Skip to main content

Full text of "A Textbook Of Geology Part I Physical Geology"

See other formats

Keep Your Card in This Pocket 

Books will be issued only on presentation of proper 
library cards. 

Unless labeled otherwise, books may be retained 
forieW weeks. Borrowers finding books marked, de- 
faced or mutilated are expected to report same at 
library desk; otherwise the last borrower will be held 
responsible for all imperfections discovered. 

The card holder is responsible for all books drawn 
on this card. 

Penalty for over-due books 2c a day plus cent of 

Lost cards and change of residence must be re- 
ported promptly. 

Public Library 

Kansas City, Mo. 

Ketp Your Card in This Pocket 

tff KOWET* ttWt 



- -> ^ ^^; Tn g*,lAiMD!]im4^^ * o 

*yy^^^ L ;^ 

%%$ m&&&. 






Physical Geology, By tlw lato L, V. Pinwwi, Third Kdi- 
tion rovimnl by Wiliam M* Agar, AHsintant ProiVnHov of 
Urology; Alan M. Batoman, PmiVnsorof Knmomir Urol 
ogy; (-art <>, Dunbar, AwHonatn Profmsor of Hmtoruml 
Geology; Richard P. Flint, AnwHUuit I'rotVwor of <Uu)t 
ogy; Adolph Knopf, Prof<wortf PhyKlrul ({eulogy; riwH- 
tr R. Ijongwtill, I'rofoKWjr of (}<>lgy, KiwiHiou Etlittnt 
by ClutKtur Jt, r^mgw^ll. ( 1 loth; byi^j 48H pagtm; JJ2iJ 


Historical Geology, By t'harUw S^htiuhort. Bocouti K<li- 
T JUwriUuu and BnlargtHl, ( 1 l<ith; (U>yi>; 7^ 
ilguroH in tttxt, 47 platen, and foiling colonul 
map of North 

Introductory Geology. By tlw latt L, V, Pi^Hon and 
(*harl<H Stdumh(*rt,. Iixduding '* l*hyioal (J^ology/' and 
an out-limt of u Hi*t<>ri(*al (<ily, n (Moth; (U>y 0; tfttf 

^5>5 ii^unw hi tixt, 26 plattm, and folding t'oloml 

;al map of North Aitwrica, 

OutHnea of Physicul Geology* rropanMl from th Thlnl 
Kdition of Part I of A Textbook of (Joology by tlw tat^ 
LOUIH V, l*irHon y and TharleB Sdmchwt, By (MuHt*ir It* 
LtngWidi, ProfoHHor of fltu>logy in Vatt* Univ^mity. Clotii; 
B by i^; 370 pagw; ^78 liguron, 

Outlines of Historical Geology, Part II of Introductory 
Utsology, By tiu* laU* L, V* I'irnsou and ('luirlos Schurh- 
crt. ("iotli; <J by U; #00 itttflun; H4 llgurtw in toxt, S4 

Mniption of fliilomnuwau, Kilnia Volnuun Hu\snii, Muy 21, 192 
flower cloud" wan Considerably nioro than a wilt 1 hirft whtn (lu 
iakciL (Miu^haru, Hilo, Hawaii,) 















W iia JAM M. A<JAH, Awsisi-ant/ Profesnor of Geology 
AbAN M, BATKMAN, Profosnor of Economic Geology 
OAHL (>. DUN BAH, Associate Professor of Historical Geology 
HiriiAUi) I (1 . KM NT, Assint.ant Profcwsor of Geology 
AOOLPU KNOPK, ProfcHsor of Physical Geology 
WH 11. LONCJWIOJUL, Professor of Geology 

Revision lOdlted by CH^BTJUH II. LONOWBLL 
I'rinting Corrected 






The textbook of geology by Pirsson and Schuchert appeared four- 
teen years ago, and the first revision of Part I was completed shortly 
before Professor Pirsson's death in 1919. Both parts of the text have 
been indorsed by numerous universities and colleges; but teachers who 
know Pirsson's volume from repeated use have come to feel the need of 
important changes in it to bring the subject matter up to date and to 
strengthen the treatment of various topics. At the request of Professor 
Schuchert, author of Part II of the textbook, six members of the De- 
partment of Geology in Yale University have cooperated in the present 
thorough revision. The revisers have attempted to preserve the ex- 
cellent balance for which the earlier editions have been commended. 
There has been no desire to make the book more elementary, but the 
aim has been to clarify the treatment so far as possible. With this end 
in view the revisers have not hesitated to recast statements, to transfer, 
add, or omit large sections, or even to rewrite entire chapters. The 
total length is essentially unchanged, although the number of pages is 
slightly increased because a uniform size of type has been adopted. 

The older editions of the book used a two-fold division into " Dy- 
namical Geology " and " Structural Geology." From their experience 
the revisers do not favor this division as a teaching device, as it necessi- 
tates a somewhat rigid order of subjects and leads to some awkwardness 
in the treatment. For example, earthquakes can be explained to best 
advantage after the student has considered the effects of crustal move- 
ments as seen in structural features; and some aspects of landscape 
development cannot be appreciated fully without knowing the nature 
and the scale of folds, faults, igneous bodies, and numerous other things 
that ordinarily are classed as structural. A proper understanding of 
geology comes only with the realization that the Earth is dynamic and 
changing. It appears that this viewpoint is presented to best advantage 
by considering the results of past activity along with the analysis of 
dynamic processes. This method gives a unified picture which is 
partially lost by any deliberate separation of the dynamic from the 
static aspects of the Earth. 

No claim is made that the order of presentation chosen for this edition 
is the best possible. Probably there will always be considerable differ- 



ence of opinion in this matter, because the subjects treated in geology 
are so interrelated that some anticipation and repetition is unavoidable 
with any order of arrangement. Of necessity the study begins with some 
consideration of minerals and rocks; but as laboratory materials are 
required, and as most instructors use some kind of laboratory manual 
dealing with these materials, no introductory chapter on minerals and 
rocks is included. Some of the common minerals are described briefly 
in an appendix, and as methods of studying the minerals are also ex- 
plained this appendix may serve as a short manual for the student 
or for the interested layman. Descriptions of the common igneous, 
sedimentary, and metamorphic rocks appear in the chapters that give 
systematic discussions of these rock groups. 

The following conspicuous changes appear in the book as revised: 
The introductory chapter is completely rewritten. Discussion of the 
larger features and relations of the Earth is transferred from former 
Chapter X to the new Chapter II. The chapter on the atmosphere is 
rearranged, with the idea of emphasizing soil formation instead of wind 
erosion; and a short discussion of weather and climate is added. The 
treatment of stream erosion is considerably amplified; the erosion cycle 
is emphasized, and new sections explain erosion in semiarid and arid 
climates. Glaciers and glaciation are discussed directly after streams, 
and ground water is placed before lakes and swamps. The former 
chapter dealing with the geologic role of life is omitted and its diverse 
subject matter appears in the treatment of weathering, swamps, and 
the ocean. At the suggestion of Professor Schuchert the chapter on the 
ocean has been expanded by the transfer of considerable material from 
Volume II of the text. The discussion of sedimentary rocks is presented 
as soon as all the agencies responsible for their formation have been con- 
sidered. Igneous rocks are discussed directly after volcanoes. Fold- 
ing, faulting, and warping are brought together in one chapter, and the 
discussion of earthquakes follows. The treatment of metamorphism 
has been changed to make it accord with present conceptions. A new 
chapter on land forms deals with the more complicated aspects of river 
history, explains the sculpturing of mountains in various stages of their 
development, and seeks to give the student some appreciation of land- 
scape as related to geology. The discussion of ore deposits has been 
simplified considerably. A geologic time table is appended because the 
revisers have found it advisable to give students a general acquaintance 
with the time scale before they begin the formal study of historical 
geology. An annotated list of references is added to each chapter for 
the use of those who may want further reading on any subject. The 
geologic map of North America issued with the older editions is omitted. 


The part performed by each of the revisers is indicated below. 

Chapters 1, 12, 13, 15, 16, by Longwell 
Chapters 2, 10, 11, 14, by Knopf 
Chapters 3, 5, 18, by Bateman 
Chapters 4, 7, 17, by Flint 

Chapter 6 and Appendix A, by Agar (now of Columbia "Univer- 
Chapters 8 and 9 and Appendix B, by Dunbar. 

As the revision progressed the work of each man had the benefit of 
criticism by his colleagues; and finally the entire revision passed through 
the hands of a single editor. 

An attempt has been made to improve the illustrations as well as the 
text. Block diagrams are used freely to explain the development of 
surface forms and to illustrate structural features. Philip B. King 
executed most of these diagrams. Many halftones have been replaced 
by new views, and the total number has been increased somewhat. 
Individual credit is given for photographs used; but the revisers wish 
to mention their special obligation to the II. S. Geological Survey, the 
U. S. Army Air Corps, and the Hawaiian Volcano Observatory. 


January 15, 1929. 
























INDEX 465 



Man has a natural curiosity about the Earth, his home. Of what 
materials is it made? When and how did it come into being, and through 
what changes has it passed? What has been the story of life on the 
Earth, and exactly what part has man himself played in the drama? 
Inquiries such as these have engaged the attention of thinking people 
from a very early period, as evidenced by the mythologies of the ancients. 
Some of the old philosophers Pythagoras, Aristotle, and others 
caught brilliant glimpses of the truth; but the science of geology, which 
strives for the full answer to questions about the Earth and the record 
of life, had its first consistent development in modern times, with the 
growth of science in general. 

Any understanding of the Earth must begin with some knowledge of 
the substances that compose it. We are permitted to see only a thin 
rind of the globe, and therefore all information about the vast interior 
portion must depend on indirect evidence. However, our most active 
interest lies in the part that can be explored directly; and this part in 
itself provides an almost limitless field of study. It is a zone composed 
of rocks and their constituent minerals. Even to the casual observer it 
is evident that these materials, exposed in cliffs or in road cuts, tunnels, 
and other artificial excavations, are highly varied in character; and 
accordingly it may appear that only a specialist can hope to gain any 
adequate acquaintance with rocks. Fortunately it has been found that 
comparatively few types are important quantitatively in the visible 
part of the Earth, and therefore any educated person may learn without 
great difficulty to recognize most of the rock masses he may see in the 
Alps, the Rocky Mountains, or elsewhere in his travels. A study of 
variations in rock types or of the more detailed features in minerals 
must of course be left to men specially equipped for the task. 

The practical value of recognizing one rock from another is readily 
apparent from the viewpoint of the professional geologist or the mining 
engineer. In the search for petroleum and for ore minerals, in the se- 



lection of sites for great dams to make storage reservoirs, or in construct- 
ing an intricate subway system for a large city, knowledge of rocks and 
their peculiarities is a vital necessity. But there is a much broader 
interest in the subject, and the general key to this interest is within easy 
reach. A visitor to the slopes of Vesuvius or of Mauna Loa sees masses 
of dark, slaggy rock. It is fairly obvious, even without actually seeing 
fluid lava, that this dark material was once liquid and flowed down the 
slopes as a red-hot stream. Continued investigation would convince 
the traveler that the fire-made or igneous rocks are common in many 
lands and have been formed at many different dates; they constitute 
an important part of the bedrock beneath us. Again, a brief examina- 
tion in parts of the high Andes or Himalayas reveals layers of compacted 
sand or mud which include an abundance of sea-shells. Phenomena of 
this kind recall the conclusion of Aristotle, " The relation of land to sea 
changes, and a place does not always remain land or sea throughout all 
time/ 7 

Thus the grouping of rock masses according to the mode and place of 
their origin is not merely a dull classification for the convenience of 
scientists; it is the first step in the fascinating game of unraveling the 
ancient history of a region. In all of its many aspects, the study or 
practise of geology recognizes this fundamental interest in past events. 
But just as the proper understanding of human history requires some 
knowledge of present day social, economic, and political conditions, so 
the deciphering of events in the geologic past is dependent on acquaint- 
ance with processes still operating on and within the Earth. Rocks 
are not inert monuments to conditions and forces that no longer exist. 
At active volcanoes we may observe all stages in the development of 
new rocks from molten material (Fig. 1). In the muds of river deltas 
and the oozes on sea floors we find the modern equivalents of ancient 
beds, now greatly distorted and eroded, which furnish the shells and other 
fossils so common in high mountains and plateaus. Rivers, glaciers, 
and other surface agencies are etching and slowly wearing away the 
lands. In some continents the land is being lifted up at a measurable 
rate, and there is evidence that some mountains are in process of active 
growth. Thus we are actual witnesses to a constant struggle between 
Titans; some that work at the surface, striving to tear down the rocky 
continents, and others within the Earth that persistently oppose the 
leveling process. Rocks are being destroyed and others are forming 
to replace them. Activities which can be seen and analyzed have been 
persistent throughout geologic time, and accordingly we may use the 
present as a key to the past. All processes now engaged in modifying 
the Earth are important in the preliminary study of geology. 


From the geologist's viewpoint, therefore, the Earth is dynamic and 
changing; not static and inert. However, since most of the changes 
are very slow as judged by human standards, it is necessary to form some 
conception of geologic time in order to appreciate the continuity of 
Earth-history. The study of geology revolutionizes ordinary notions 
of time, just as a moderate knowledge of astronomy gives a new vision of 
space. Our solar system has grand dimensions, and yet in its entirety 
it is but a point by comparison with the stupendous diameter of the 

Pig. 1. Igneous rock in the making. The dark-colored rock is solidified lava; the 
white band is a stream of fluid lava, flowing toward the observer. (The Alika flow, 1919. 
Hawaiian Volcano Observatory.) 

starry galaxy. Similarly we think of the earliest human records as 
very ancient; but in a geological sense the first appearance of man is a 
modern event. The oldest relics of primitive men are found only in the 
superficial soil or other rock debris formed in the latest geologic epochs. 
In all the time that has elapsed since the oldest civilizations existed, the 
major landscape features of the Earth have remained essentially un- 
changed; but we know from the immensely longer geologic record that 
generations of mountains were made and worn away, before any evi- 
dence of man appeared, by the same deliberate forces now at work. 
The length of time required for such transformations has been almost 
inconceivably great. How long has it taken for the Colorado River 
to excavate the Grand Canyon? We are appalled at the realization 


that the slow action of water has carried away so many cubic miles of 
solid rock. Nevertheless the Grand Canyon is a youthful feature. 
The general record of earlier events is written plainly in the rocks tra- 
versed by the trail in climbing the vertical mile from the inner gorge to 
the outer rim (Fig. 2). Mountains were made and worn down; then 
the land was submerged beneath the sea for long ages, and continued to 
sink slowly while limy ooze, mud, and sand built up a deposit thousands 


Fig. 2. A minor canyon tributary to the Grand Canyon. Colorado River shows at 
lower right. The horizontal layers of rock represent accumulations on sea floors during 
long geologic periods. The cutting of the canyon by stream erosion is a recent event 
geologically, although it required a long time measured by human standards. (U. S. 
Army Air Corps.) 

of feet thick; this loose material was converted into firm rock; and 
finally the rising of the wide plateau region high above the sea permitted 
the cutting of the canyon. Minimum estimates fix the age of the low- 
est rocks in the gorge at hundreds of millions of years; and yet these 
rocks may be youthful in comparison with the total age of the Earth. 
The closest study of the geologic record has revealed " no traces of a 
beginning, no prospect of an end." 

This broad, philosophical aspect of geology rests on a secure foundation 
only because of patient effort by generations of workers in all lands. 


Prior to the nineteenth century many natural philosophers formed their 
ideas on geological subjects largely by deductive reasoning, in ignorance 
of the facts to be learned by observation in the field. Gradually it 
came to be realized that inductive study, based on all the facts obtain- 
able, is essential for any safe conclusions. A vast array of field evidence 
has been accumulated. The great mountain ranges have been fruitful 
fields for geologic investigation, since they furnish the finest exposures 
of rock formations. But no sources of information are neglected. 
Geologic features everywhere are examined closely in the field and repre- 
sented accurately on maps. Mines give opportunity to explore beneath 
the surface, and deep wells drilled for water or oil yield valuable data. 
Modern instruments devised to record earthquake waves and to measure 
the value of gravity have made it possible to draw important conclusions 
regarding the invisible interior. Slowly but surely the Earth is giving 
up many of its secrets to the inquisitiveness of man. 

Of necessity a field so broad as the study of the whole Earth calls for 
a division of labor among specialists. Some workers give their principal 
attention to the rock formations laid down in former seas, lakes, and 
rivers; some to the volcanic and other igneous rocks; some to the mineral 
veins and other deposits of economic value. Other subjects 'of special 
study are the deformation of rocks by folding and fracturing; the various 
land forms sculptured by surface agencies; the fossils entombed .in 
rocks; and the minerals that make up rocks of all kinds. All groups of 
investigators recognize unsolved problems, and by their united efforts 
continue to wic^en the frontiers of the science as a whole. Like any 
other growing subject, geology extends beyond the lighted zone of 
proved fact. It has also a large twilight zone of inference and proba- 
bility into- which the full light of investigation continues to spread slowly; 
and beyond this a region of shadow and complete darkness, relieved only 
by scattered flashes of speculation. The speculative side of the subject 
is in some respects the most fascinating, -but it may also be dangerous 
to the uninitiated. A comprehensive discussion of geology takes ac- 
count of fact, inference, and speculation, but distinguishes carefully 
between them. Careless broadcasting of hypotheses as if they were 
proved facts has given rise to numerous popular misconceptions about the 
Earth and its life. 

The study of geology begins logically with an introduction to the 
important kinds of rocks and minerals. It proceeds to an examination 
of the forces that act on the outer part of the Earth, and the changes 
produced by these dynamic agencies. Finally, the keys provided by 
this preliminary study are used to explore the long geologic record con- 
tained in the rocks. 


The Earth and Its Neighbors. The Earth is one of a group of 
planets that revolve around a common central orb the sun. Some 
of these, like Jupiter, are much larger than the Earth; some like the 
asteroids, or minor planets, are much smaller; some are much nearer the 

sun, others farther away. The 
group has nearly a common plane 
of revolution about the sun, as 
suggested in Fig. 3, and this fact is 
thought to have an important bear- 
ing on the origin of the solar sys- 
tem. ' The Earth and other planets 

Fig. 3 Planets revolve about the sun in were k orn O f h e sun . according to 
nearly one plane, as suggested by the orbits , 1,1 c 

of three of them. the modern theory of cosmogony, 

which we owe to the American ge- 
ologist, Chamberlin, the substance of the planets was expelled from the 
sun under the influence of the disruptive pull of a passing star several 
times larger than itself; but how this ejected matter was aggregated to 
form the planets is a problem on which ideas are still far apart. The 
common plane of revolution of the sun's satellites is possibly an inheri- 
tance from the biparental origin of the system. The birth time of the 
Earth was probably 2000 million years ago, and since that time the Earth 
has been gradually acquiring its present character. A vast span of time 
has thus elapsed, and that portion the record of which is written in the 
rocks is referred to as " geologic time." 

The sun is called by astronomers a G-type dwarf star, by which they 
mean that it has already lived three-fourths of its life as a self-luminous 
body. Its surface temperature is approximately 6000 C. as determined 
observationally; the temperature at its center is estimated to be 
40,000,000. The geologic record shows that the sun has been supplying 
light and heat to the Earth at an approximately uniform rate during 
hundreds of millions of years. Up to the early part of the present 
century the other sciences could not adequately explain how the sun 
had been able to maintain this prodigal expenditure of energy. It is 
now considered probable that its fires are stoked either by utilizing sub- 



atomic energy or by the conversion of matter into radiant energy 
the annihilation of matter. According to Eddington, the transmuta- 
tion of the elements would keep the sun going 10,000 million years, but 
the annihilation of matter would give " a very ample time scale. 77 

The mass of the sun is 332,000 times that of the Earth. Even counting 
in the masses of the other planets, practically all of the mass of the solar 
system is in the sun. Therefore the sun is the overwhelmingly dominant 
member of the system. Carrying with it the Earth and the other planets, 
the sun is travelling through space at the rate of 12 miles a second. 

The path of the Earth about the sun is not a circle, but an ellipse, 
one of whose foci is the sun. The deviation of the ellipse from a circle, 
however, is relatively small; the average distance of the Earth from the 
sun is nearly 93 million miles. In consequence of its elliptic orbit, the 
Earth is 3 million miles nearer the sun on January 1 than it is on July 1. 
This fact causes the summers to be somewhat cooler and the winters 
somewhat warmer in the northern hemisphere than they would other- 
wise be. At the present time summer, as measured from the vernal 
equinox on March 21 to the autumnal equinox on September 23, is 7 
days longer than winter. In the southern hemisphere, however, winter 
is the longer season. Owing to the precession of the equinoxes, this 
condition will be reversed between the two hemispheres 10,500 years 
hence. Because of the perturbing effect of the other planets, the ec- 
centricity of the Earth's orbit varies during a period of 500,000 years, 
and at maximum eccentricity the Earth is 13 million miles nearer the 
sun in summer than in winter, thereby causing short hot summers and 
long cold winters. At that time the winters will be 35 days longer than 
summer. As we shall see later, these astronomic conditions have been 
held by some to be sufficient to produce great climatic changes, to de- 
termine cycles of sedimentation, and even to bring about the several 
glacial epochs that have affected our planet during its long career. 

Besides revolving around the sun, the Earth is spinning on its polar 
axis, each rotation in 24 hours giving rise to day and night. The axis of 
rotation is not perpendicular to the plane of the Earth's orbit but in- 
clined to it at an angle of 66^, and this inclination gives rise to the 
seasons, summer and winter, alternately in the northern and southern 
hemispheres according as the axis is pointed toward the sun or away 
from it. 

The Earth, as we have seen, is a very insignificant fraction of the 
solar system, and for that matter the solar system itself is but an in- 
significant fraction of the universe. The sun is a modest star in a stellar 
system whose members are numbered in thousands of millions. At 
immense distances beyond this system our " universe, " as it is cur- 


rently called in astronomy are other gigantic systems. According 
to Sir J. H. Jeans, " the farthest astronomical objects whose distances 
are known are so remote that their light takes over 100 million years to 
reach us." Nevertheless, throughout this vast extent, with its island- 
universes containing myriads of stars in various stages of development, 
the same general physical laws that we know on the Earth appear to 
govern. Gravity operates in the same manner; light is transmitted 
everywhere by the same kind of vibrations; the spectroscope tells us 
that the same chemical elements occur in distant stars as on our Earth. 
Moreover, the meteorites that our planet gathers in its journey through 
space and which appear to be the disrupted fragments of former worlds 
are made of substances identical with those found on the Earth. Con- 
sequently there appears to be a unity of law and a uniformity of material 
throughout space, and we feel justified in assuming that facts and reason- 
ing derived by astronomical study of the other heavenly bodies may also 
be applied in our study of the Earth. 

Form of the Earth. The Earth is not a true sphere but a spheroid 
flattened at the poles, so that the axis on which it rotates is slightly 
shorter than the equatorial diameter. The polar flattening (oblate- 
ness) of the Earth and the other planets is fully explained by the prin- 
ciples of celestial mechanics. It depends on the centrifugal force due 
to rotation and on the internal constitution of the planet. By internal 
constitution is meant the composition, density, and distribution of the 
density within the planet. 

It is now accepted that the rate of rotation is gradually decreasing, 
or in other words that the day is growing longer, at the rate of j-^ 
second per century. The reason advanced for this diminution is that 
the fluid friction that is caused by tidal turbulence in the shallow seas 
on the borders of the continents uses up energy, and as this energy is 
derived chiefly from the Earth's energy of rotation it therefore retards 
the rate of rotation. Two-thirds of the frictional loss occurs in Bering 
Sea, a shallow body of water having strong tidal currents. As the size, 
depth, and number of such seas has varied greatly throughout geologic 
time, in fact a large part of Historical Geology is the record of the ex- 
pansion and contraction of such seas, it is plain that the rate of retarda- 
tion cannot have been constant. 

On the other hand, it is probable, though not proved, that the Earth 
has been contracting through loss of heat; because of the resultant 
shrinkage in size it would rotate faster. This increase might well coun- 
terbalance the retarding effect of the tidal friction. The problem of 
the net increase or decrease of the Earth's rotation during geologic time 
must therefore be set down as unsolved. 



General Features. The irregularities of the Earth's surface, or its 
relief, divide naturally into major and minor groups. The major relief 
features are the continents and the ocean basins. The continents 
are regarded as being partly submerged beneath the oceans, and the 
submerged portions are known as the continental shelves. The minor 
relief features of the continents are the mountains as opposed to valle3 r s 
and basins; and the minor relief features of the ocean basins are islands 
and submarine ridges as contrasted with the profound troughs or deeps 
in the ocean floor. 

The average height of all the lands above sea-level is 2300 feet. 
North America averages about 2000 feet. The average depth of the 
oceans is about 13,000 feet. The highest elevation of the land, Mt. 
Everest in the Himalayas, is 29,000 feet; the greatest known depth in 
the ocean, in the Pacific, is 35,400 feet. Thus the greatest difference in 
relief exceeds 64,000 feet, or more than 12 miles. Relative to its size, 

Fig. 4. Mt. Everest in comparison to the size of the globe. Part of the arc of a globe 
with radius of one foot is shown ; on this Mt. Everest, E, would be about ^ of an inch high. 
D, in a similar way, shows the greatest depth of the ocean. 

however, the globe has an extremely small relief, and it is therefore com- 
paratively smooth; Fig. 4 shows its greatest roughness. 

The relief features of the land are the plains, such as the Atlantic 
Coastal Plain; plateaus, such as that of the Colorado; and mountains, 
like the Appalachians extending from Cane da to Alabama. In regard 
to the grouping of the relief forms of the Earth, certain facts are of in- 
terest and importance. The continents as a rule consist of basins that 
are bordered by mountain chains along the coastal rims, whereas the 
ocean basins generally have the reverse arrangement, the deeps being 
near the continents and the submarine ridges, or upswells of the bottom, 
being in mid-ocean. Some of the highest and most important ranges on 
the edges of the continents border the greatest deeps in the ocean 
floor, as for instance the Andes in South America and the partly sub- 
merged mountain chain that forms the Japanese islands and is the real 
eastern border of the continent of Asia; close to these ranges the ocean 
floor descends to great depths. It is not meant to imply, however, that 
mountains occur only at the continental edges, for they may extend in 
a wide zone far into the interior, as in western North America, or form 
systems crossing a continental mass, as in Asia. 


Character of North America. North America is regarded as the 
most typical of the continents. It is bordered by mountainous tracts 
on either side and contains the great basin of the Mississippi and its 
tributaries in the interior. The following broad features of the conti- 
nent and especially of the United States will enter into many of the dis- 
cussions of its geology. 

On the east and south the continental shelf rises from the sea as the 
Atlantic Coastal Plain, and this plain extends to the base of a rugged 
mountainous tract of country, which stretches from Canada into 
Alabama and is known as the Appalachian Highlands; it includes the 
Appalachian Mountains. This mountainous belt gives way to the vast 
basin whose higher western part forms the Great Plains. The lower 
part of the basin is the Central Lowland, from which the Mississippi 
descends through the Gulf Coastal Plain. On the west the Great 
Plains abut upon the long series of north and south ranges that form the 
backbone of the continent and are grouped under the name of Rocky 
Mountain System. Between this system and the Pacific Mountain 
System, which makes the western rim of the continent, are the Inter- 
montane Plateaus, consisting from north to south of the Columbia 
plateau, the Great Basin, and the Colorado plateaus. The Pacific 
Mountain System consists of the Sierra Nevada, Coast Ranges, and 
Cascade Range with the Pacific border lands. These relations and other 
minor ones can be seen in Fig. 5. 

The divisions into which the United States and southern Canada are 
divided on the basis of physical features that have a common geologic 
history are shown on the map, Fig. 5. These physiographic regions are 
classified into major divisions, shown by letters; and minor provinces, 
indicated by the subscript numbers. Their names are given in the 
following list: 

Major Divisions Provinces 

. T ,. f AI, Laurent ian Plateau. 

A. Laurentian Upland < . . TT , , 

[ Aa, Superior Upland. 

D AI 4.* -ni f BI, Continental Shelf (submerged). 

B. Atlantic Plain 1 -D ^ A i 

[ B 2 , Coastal Plain. 

C. Appalachian Highlands . 

Ci, Piedmont Province. 

C2, Blue Ridge Province. 

Cs, Appalachian Valley and Ridge Province. 

C 4 , St. Lawrence Valley. 

C 5 , Appalachian Plateaus. 

C 6 , New England Province. 

Cy, Adirondack Mountains. 





D. Interior Plains . 

E. Interior Highlands . 

F. Rocky Mountain System . . , 

G. Intermontane Plateaus . 

H. Pacific Mountain System . 

Di, Interior Low Plateaus. 
Do, Central Lowland. 
D 3 , High Plains. 

Ei, Ozark Plateaus. 
E 2 , Ouachita Province. 

FI, Southern Rocky Mountains. 

Fo, Wyoming Basin. 

F 3? Northern Rocky Mountains. 

Gi, Columbia Plateau. 
G2, Colorado Plateaus. 
G 3 , Basin and Range Province. 

H 1; Sierra-Cascade Mountains. 
H 2 , Pacific Border Province. 
Hs, Lower California Province. 


It is chiefly the outer zone of the Earth about which we have extensive 
and positive information. This outer zone is known as the crust; it 
is built of rocks, which are discontinuously covered by a thin mantle of 
soil. Because " crust " is thought by some to connote that the Earth 
has a liquid interior, the term lithosphere has been coined to avoid this 
implication, but we shall use the simpler term preferably. It is upon 
the crust that we live and exert our activities; we penetrate into it for 
fuels of various kinds, for metals, water, building material, and other 
mineral resources, all of which are essential to the physical side of modern 
civilization. A thorough knowledge of the component parts of the 
Earth's crust and its structure is consequently of the highest importance. 
The component parts of the crust are rocks, and we shall begin our in- 
quiry by a study of the different kinds of rocks and the varied modes 
in which they occur. 

Definition and Classification of Rocks. The word rock, in the 
language of geology, means the material that composes one of the in- 
dividual parts of the Earth's outer shell. According to their mode of 
origin, rocks are divided into three main groups: the igneous rocks, 
made by the solidification of molten material; the sedimentary, or 
bedded rocks, formed from sediments that were deposited chiefly by 
water (and to some extent by air and ice) ; and the metamorphic rocks, 
formed by certain processes acting within the Earth's crust on pre- 
existing rocks and partly or wholly destroying their original characters 
and producing new ones, so that the resultant rocks are best considered 
as constituting a separate group. 



Thus we have three groups: 

I. Igneous Rocks consolidated molten masses. 
II. Sedimentary Rocks formed from sediments deposited by 

water, air, or ice. 

III. Metamorphic Rocks secondary, derived from preexisting 

Three-fourths of the land area of the globe is underlain by sedimentary 

y^rocks and the other fourth by igneous and metamorphic rocks. Al- 

though the sedimentary rocks thus preponderate in the visible part of 

x\the crust, they are essentially a veneer, a mile or less thick on the aver- 

jage. The foundation rock of the continents is largely igneous rock 

N (granite), which is probably between 10 and 20 kilometers thick. The 

^detailed discussion of the constitution of the crust is reserved for Chap- 

ter XV. 



1. The Fundamentals of Astronomy; by S. A. Mitchell and C. G. Abbot. 307 
pages, D. Van Nostrand Co., New York, 1927. 

2. An Introduction to Oceanography, with special reference to geography and 
geophysics; by James Johnstone. 368 pages. 2nd edition, Hoddar and Stoughton, 
Limited, London, 1928. 



General Functions of the Atmosphere. The atmosphere is directly 
essential to life on the Earth, and in addition it enables the sun and the 
rain to bring about life-giving changes on the Earth's surface. With- 
out the atmosphere the Earth would be boiling hot by day or freezing 
cold by night, and devoid of life, as the moon is supposed to be. There 
would be neither wind nor rain nor bodies of fresh water. The lands 
would be rugged and desolate. Winding streams and soft land- 
scapes mantled with colored soils and vegetation, and other vistas 
pleasing to the eye, would be lacking. For all these things, as we shall 
see/ result directly from the presence of the atmosphere. Due to the 
rotation of the Earth and unequal heating by the sun, movements are 
set up in the atmosphere that account in part for our climate and weather. 
The wind itself builds up structures such as sand dunes, or abrades the 
rocky surface by the sand it carries along. But far more important are 
the chemical action of air and moisture, and the mechanical effects of 
changes in temperature, or frost action, aided by plant and animal life, 
that cause rocks to change into soil. This is called weathering, and the 
soil so formed may be swept away by the rain that falls from the at- 
mosphere, and as more soil is formed from the rocks beneath, it in turn 
is carried off, and so the surface is gradually wasted away. This general 
process, involving the wearing away of the land, we call erosion. In 
this brief picture the importance of the atmosphere is evident, and we 
will now consider some of these factors in more detail. 

Character of the Atmosphere. The outer gaseous envelope of our 
globe is known to extend at least 200 miles above the Earth, since meteors 
heated to luminosity by friction with the outer air have been observed 
at this height. It must extend considerably higher. But even at a 
height of 50 miles it is extremely attenuated, and at 19,000 feet, the 
height of Mount St. Elias, its density is only half that at sea level. Its 
weight is estimated to be 1/1,200,000 of that of the Earth, or about 
5 quadrillion tons. This weight exerts a pressure of nearly 15 pounds 
to the square inch, or a ton to the square foot, at sea level; but since 
gases transmit pressure equally in all directions, the pressure beneath 
any object is as great as that above, and hence the effect of the atmos- 
phere's weight is not noticed. 



The atmosphere consists of three chief constituents in the following 
quantities by weight: nitrogen, about 75 per cent; oxygen, about 23 
per cent: and argon, about 1.4 per cent. There are also present in small 
quantities the inert gases krypton, xenon, helium, and neon, but these 
are of no importance geologically. In addition, water vapor (average 
is 1.2 per cent at Earth's surface), carbon dioxide, hydrogen dioxide, 
ozone, ammonia, and traces of hydrogen, sulphur, organic matter, and 
suspended solids, are present in varying quantities. But of all these 
constituents the only ones that play an outstanding part in geologic 
processes are oxygen, water vapor, and carbon dioxide. 

The oxygen, in addition to its obvious function of sustaining life, also 
enters into chemical changes that are important geologically. Wherever 
animals breathe or fire burns, oxygen is withdrawn from the air and 
locked up in compounds such as carbon dioxide. But growing plants, 
on the other hand, liberate some oxygen from carbon dioxide. Also 
the weathering of rocks involves the production of iron oxides such as 
limonite, in which oxygen is absorbed from the air. Thus the atmos- 
phere is thought to be undergoing a slow net depletion of its oxygen 
content, though so slow that it is not detectable by chemical tests. 

The carbon dioxide content in the atmosphere is relatively constant 
at 3 parts in 10,000 by volume. Though the quantity is small, its geo- 
logical importance is great. Carbon dioxide is being added to the at- 
mosphere from volcanic and mineral spring emanations, the combustion 
of fuels, the respiration of animals, and the decay of organic matter. 
It has been estimated that the consumption of coal annually returns to 
the atmosphere about 1/1000 of its present content, so that in 1000 
years the amount of carbon dioxide in the atmosphere would be about 
doubled. But other operations are as continually diminishing the 
carbon dioxide content, notably its extraction from the air by growing 
plants, and its great consumption in the weathering of rocks. 

The carbon dioxide and water vapor of the atmosphere also exert a 
beneficial effect on the Earth by causing part of the solar heat to be 
retained. Thus they help make the Earth hospitable to living things, 
and they aid chemical reactions that are important in the formation and 
the destruction of rocks. A comparatively slight increase in the carbon 
dioxide content of the atmosphere would bring a warmer climate, and a 
decrease of only a few per cent would lower temperatures appreciably. 


It is obvious that rock weathering and erosion will proceed in a differ- 
ent manner in regions where rain falls intermittently throughout the year 
than in places where rainfall is seasonal; also the effect of these processes 


will be unlike in an arid climate and a humid climate. Consequently, 
a knowledge of climate and weather is essential to a proper understanding 
of the character and distribution of the various climatic regions, and of 
the geologic processes that operate within them to produce far-reaching 
changes on the Earth's surface. And this involves consideration of 
atmospheric movements, moisture, temperature, and pressure. 

Movements of the Atmosphere. Air heated in the equatorial re- 
gions expands, rises, flows poleward, and settles down some distance 
north and south of the equator, and then moves over the surface toward 
the equator. The colder polar air tends to move equatorward. This 
circulation resembles that of the hot water in a house-heating system, 
and like the latter is induced and maintained by temperature differ- 
ences. A great atmospheric circulation is thus set up. If the Earth 
did not rotate, and were smooth, and if its temperature varied uniformly 
from equator to poles, just such a simple atmospheric circulation would 
prevail. But this ideal simplicity does not exist. The rotation of the 
Earth from west to east throws this otherwise simple circulation very 
much askew, so that the air moving toward the poles is deflected east- 
ward and that moving toward the equator is deflected westward. There 
are thus set up in the lower part of the atmosphere belts of prevailing 
or planetary winds (Fig. 6). 

Those in the equatorial regions, roughly between latitude 30 N. and 
S., are the easterlies or familiar trade winds, and those in the middle 
latitudes are the westerlies. Where the trade winds from the northern 
and southern hemispheres meet near the equator, there is a belt of up- 
rising air attended by cloudiness, high humidity, and calms, known as the 
doldrums. And between the trade winds and the westerlies there are 
the other belts of calms or light variable winds where the barometric 
pressure is high, the air descending, humidity low, and skies clear, 
known as the horse latitudes. Other movements due to differences in 
summer and winter heating over land and sea give rise to summer and 
winter monsoons, of so much importance to the climates of India, China, 
Australia, parts of Africa, and Texas. All of these are fundamental air 
movements that persist with regularity on different areas of the Earth's 

But in the middle latitudes there are, in addition, minor or secondary 
movements known as cyclonic storms, which affect animal and plant 
life from day to day and give rise to our weather. The westerlies of 
these latitudes are not simple prevailing winds always from the west. 
Here the equatorial and polar currents intermingle and, due in part to 
imperfectly understood effects of the upper atmosphere, and in part to 
unequal heating and cooling, circular air movements called cyclones 



(low pressure areas) and anticyclones (high pressure areas) are set up. 
These are essentially thin, flat discs* of moving air from 500 to 1000 
miles across, that sweep easterly at an average velocity of 500 to 800 
miles a day. Within these discs, the movement may be pictured as 
whirlwinds in which, in each of the low pressure areas (the Lows), 

Fig. 6. To show the direction of planetary winds on the surface and in the upper 
atmosphere. (Modified after Ferrel, and Tarr and Martin.) 

a rising spiral of warm air rushes anticlockwise toward the center in 
response to the low pressure ; whereas in each of the high pressure areas 
(the Highs) a descending column of cold air spirals toward the outside 
in a clockwise direction, being pushed out, as it were, by the high pres- 
sure. In the southern hemisphere the directions are reversed. About 
three Highs and three Lows may be expected to pass from west to east 
over any point in central and northeastern United States every two 
weeks. The direction of the wind at a given place at any particular 
time will depend upon which part of a Low or a High covers it; also, 
as the Low or High moves farther east, the direction of the wind at a 
given place will change. Thus the weather in the middle latitudes is 
continuously changing. 


Atmospheric Moisture. Water in the form of vapor is always 
present in the atmosphere, though its amount varies greatly from time 
to time and from place to place. When the atmosphere carries its full 
capacity it is said to be saturated, but this condition depends largely upon 
temperature, since warm air can contain more moisture than cold air. 
If air is only one-half saturated for a given temperature, it is said to have 
a relative humidity of 50. The average relative humidity over the oceans 
is about 85; over the lands it is considerably less. If air with a relative 
humidity of, say, 60 is cooled steadily, though there is no change in the 
actual amount of water present, the relative humidity will increase up 
to 100, or saturation, and further cooling will cause condensation. 
Upon this simple principle depends the control of precipitation, whether 
in the form of dew, cloud, fog, rain, or snow. 

In nature there are several possible causesjor cooling of moisture- 
laden air, such as radiation, moving of cooler air to it, or rising of the 
moist air into higher altitudes where expansion causes lower tempera- 
tures. The latter cause, depending on the simple principle of expan- 
sion, is the commonest. No surprise can be felt, therefore, at the 
heavy rainfall on mountains that lie in the path of air currents moving 
from the sea, for we know that .high ground forces the air to move 
upward to cooler altitudes, where further cooling by expansion aids 
condensation into clouds and rain or snow. The same thing occurs 
when warm, moisture-laden air moves upward in the center of a low- 
pressure area, and that is why cloudy weather or precipitation accom- 
panies Lows. 

Conversely, if the air is warmed, its capacity to absorb and hold 
moisture is increased and it becomes undersaturated. Consequently, 
the cold air that falls in the center of a High becomes warmed by com- 
pression and is a drying wind from which no precipitation occurs. 
Similarly, air that falls on the lee side of a mountain range becomes 
warmed in its descent; if the fall be rapid and great, the skies are clear, 
the air is warmed, and any surface water in its path is rapidly absorbed. 
In the Rocky Mountains such a hot, drying wind is called a Chinook; 
in the Alps it is called the Fohn, and there the peasants welcome it be- 
cause it melts the snows to water the pastures and it ripens the grapes 
and grain. 

Atmospheric Temperatures. The atmosphere receives most of its 
heat from the sun and, as pointed out above, this solar heat is retained 
largely through the action of carbon dioxide and moisture in the air. 
More heat is absorbed from the direct rays of the sun near the equator 
than from the slanting rays near the poles, and consequently the tem- 
perature, generally speaking, decreases from the equator to the poles. 



But the uniformity of this arrangement is disturbed by ocean currents 
and the disposition of land and sea. Land absorbs more of the sun's 
heat than does water, and gives it off more readily; consequently it 
becomes warmer by day and cooler by night than does water. Water 
retains more of the heat it receives and tends to store it up until it is 
in part distributed by currents. For these reasons land climates are 
more variable than oceanic climates, and maritime regions are milder 
in winter and cooler in summer than continental areas. In the middle 
latitudes, local changes in temperature are sudden and pronounced, 
owing to the alternation of Highs and Lows. 

Weather. In middle latitudes the procession of Highs and Lows 
brings changing winds and temperature, and clear or rainy spells. Thus 

Fig. 7. Weather map of the United States for Jan. 5, 1929. Shows high and low 
pressure areas, and a storm in the Mississippi Valley. 

the weather moves and varies. In North America a low pressure area 
moving generally eastward may in its broad reach cover a considerable 
part of the eastern United States and southern Canada (Fig. 7). Rains 
occur on the eastern and southern side of the Low because the winds 
are warm and moisture-laden, after their sweep over sea waters to the 
south and east. When they ascend in the cyclonic center, cooling takes 
place and condensation soon occurs, so that cloudy or rainy weather 
accompanies the Lows. This eastward movement of Lows, then, gives 


a generous rainfall to the eastern part of the continent. Occasionally 
the Low may remain nearly stationary, and then rains may be so con- 
tinuous that floods result. 

As a Low passes and a High follows, the warm winds from the south- 
east and south shift to the west and northwest and the temperature 
falls rapidly. This causes the cool spells of summer and cold waves of 
winter. The descending cool air is undersaturated ; drying winds and 
clear skies result. 

Meteorological observations gathered from scattered points make it 
possible for weather maps such as Fig. 7 to be plotted daily, and by use 
of the principles just discussed and with the understanding that weather 
travels, reliable weather forecasts can be made 24 to 36 hours in advance. 

Climate. Climate has been defined as the " sum total of meteoro- 
logical conditions that constitute the average state of the atmosphere 
at a given point on the Earth's surf ace. " 

If the Earth had a smooth, homogeneous surface and no atmosphere, 
-there would be simple solar climates whose distribution would corre- 
spond with latitudes. While it is true that latitude is the most impor- 
tant single factor controlling climate in so far as temperature is con- 
cerned, there are other factors that profoundly modify it. Some of these 
have already been mentioned, such as proximity to large bodies of water 
or ocean currents, prevailing or seasonal winds, mountain ranges, cy- 
clonic storms, and altitude. 

Away from the equatorial belt the ocean currents are the chief con- 
trolling factor in the climates of lands that border the seas. For ex- 
ample, compare the equable climate of the British Isles, tempered by 
winds from the warm Gulf Stream, with that of Labrador at the same 
latitude, washed by the chilly Labrador current. Or compare the mild 
climate of the coast of British Columbia and Alaska, where the influence 
of the Japan current is felt, with the harsh climate of the Siberian coast 
directly opposite. This contrast between east and west coasts is pres- 
ent nearly everywhere. A region in which the winds blow from an 
adjacent ocean has, for its latitude, mild winters and cool summers and 
small daily variation in temperature, features typical of a maritime cli- 
mate. But it will be noted that in high latitudes true maritime climates 
are restricted to the west coasts of continents, whereas the east coasts, at 
corresponding latitudes, are influenced by continental conditions. 

In the regions of the monsoons, notably in Asia, the year is clearly 
divided into a dry season and a period of torrential rains, when moisture- 
laden winds blow toward the lands. If these winds are intercepted by 
mountains, there is a particularly heavy rainfall for the reasons previ- 
ously discussed. 



Continental climates are marked by great variation in temperature, 
rainfall, and barometric pressure. In the United States, for example, 
a thin maritime belt (Fig. 8) fringes the northwest coast where extremes 
in temperature are few and rainfall is plentiful. The prevailing westerly 
winds carrying moisture from the Pacific Ocean are forced high by the 
coastal ranges, whose western slopes are thus well watered. Where 
the dry winds drop suddenly on the eastern lee slopes, there are arid and 
semiarid belts where temperature changes from night to clay are pro- 
nounced. The contrast with the coastal belt is striking. Throughout 

Fig. 8. Rainfall map of the United States. (U. S. Geol. Surv.) 

the Great Basin region arid or semiarid climates prevail and the daily 
and annual variation in temperature is extreme. The climate of the 
Mississippi drainage basin and the Eastern States is controlled largely 
by latitude, cyclonic storms, and proximity to the Atlantic Ocean. 
Sudden changes in temperature are frequent, and clear and rainy weather 
alternate throughout the year. The New England States, over which 
most of the cyclonic storm paths converge, have particularly change- 
able weather, with fairly evenly distributed precipitation throughout 
the year, and pronounced temperature changes. Warm and cold spells 
alternate with surprising suddenness, even in midwinter, due to a High 
following a Low. This is typical of a continental climate in middle 
latitudes. An extreme example of the continental climate is central 
Asia, which is not only remote from the oceans, but is also cut off from 


ocean winds by mountain barriers. In consequence of these various 
factors, different climatic provinces are produced throughout the world. 

Factors that Change Climates. From what has been said above 
it is clear that if the controls of climate be changed, then a different 
climate for a given place will result. Examples are changes in the dis- 
tribution of land and water, with a shifting of the ocean currents; 
the formation of a new mountain range on the land; changes in solar 
radiation, or other astronomic changes; and changes in the carbon diox- 
ide content of the atmosphere, or in the atmospheric circulation. The 
studies in geology to follow will teach us that in the past just such changes 
of climatic controls have taken place. 

Climates of the Past. The records of geology show that there have 
been profound changes of climate in the course of the Earth's history, 
and such changes are probably still in progress. Fruitful lands of today 
were once barren deserts; corals and subtropical plants thrived in what 
are now cold northern regions. And in the period just preceding our 
own, a widespread glacial climate prevailed over the regions of high 
latitudes, and ice sheets spread over the lands. Still earlier, glaciation 
spread over subtropical parts of Africa, Australia and India. Even 
within historic times changes in increased rainfall or in progressive 
desiccation have been noted. An example of the latter is to be seen in 
the part of northern Africa that was once the granary of Rome. 

These climates of the past, in their changing character throughout 
the ages, have influenced notably the development of life upon the 
Earth, and have affected rock weathering and other geologic processes 
that have operated within their confines. 

X , 


The records of geology teach us that the Earth is constantly changing, 
and nowhere is this so evident as on the Earth's surface where the atmos- 
phere and the lithosphere meet. Even the hardest rock cannot with- 
stand the slow unrelenting attack of the atmosphere; it will crumble 
away to fragments or soil. 

The attack of the atmosphere upon the rocks is twofold mechani- 
cal and chemical. The first produces disintegration of the rocks by me- 
chanical agencies such as frost action, changes in temperature, wind or 
rain. The chemical attack results in decomposition of the rocks, and a 
certain amount of this is essential to form good soil. Both kinds of 
destruction usually operate together and each is aided somewhat by the 
work of plants and animals. The dominance of one or the other de- 
pends to a large degree upon the nature of the climatic provinces in 



which they operate. Together they constitute a rather complex set 
of processes called weathering. 

Weathering is aided greatly by the shattered character of the bedrock, 
which is penetrated in all directions by cracks and fissures, some great, 
some small (Fig. 9). Even the mineral grains of the rocks are more or 
less filled with cracks. These openings are potent factors in promoting 

v *'"* ' : '*-lM 

Fig. 9. Shattered nature of the bedrock which facilitates disintegration and decom- 
position by allowing ready ingress of water, air, and plant roots. 

disintegration and decomposition, for they allow ready entrance of air 
and water and plant roots into the rock. 

What these various agencies are that together produce weathering it 
will now be our purpose to consider. 


The Expansive Force of Freezing Water. Water pipes that burst 
upon freezing are a familiar manifestation of the expansive force of 
freezing water a force that exceeds 2QOO pounds per square inch 
In the days of muzzle-loading cannon, captured enemy guns were fre- 
quently disposed of by being filled with water and left to freeze, by which 
means they were split from end to end. The effect upon rocks is similar; 
water trickles into the numerous crevices, joints, and pores (Fig. 9), 
and upon freezing, pries off small pieces or large blocks. The latter 
are in turn similarly broken into smaller pieces. In the same manner 
porous sandstones may be completely reduced to sand. This action goes 
on whenever water freezes in a confined space. Naturally it is most 
pronounced where freezing and thawing occur frequently, as in most 


cemperate regions in the fall and spring. This is particularly the case 
3n high mountains where, almost regardless of latitude, thawing by day 
and freezing by night may go on over a considerable part of the year. 
Anyone who has camped on a steep mountain on a frosty evening will 
aever forget the startling noise of falling rocks dislodged by this process. 
Debris formed in this way contributes to the formation of talus slopes 
'Figs. 10, 14), containing many millions of tons of broken rock, which 
:lank the lower mountain slopes, particularly in high latitudes. Those 
rfio climb mountains are familiar also with the masses of broken rock 

Fig. 10. Rock disintegration and weathering in high, altitudes with formation of 
long talus slopes of slide rock. Mt. Sneffels, Colorado. (U. S. Geol. Surv.) 

Dn the tops of ridges. These are impressive illustrations of the effective- 
less of frost action on larger topographic features. The results of this 
process are also clearly evident on many buildings and tombstones 
throughout the northern States, and should be taken into account in 
ihe proper selection of building stones. Had the Sphinx been set up in 
New England instead of in the warm, dry climate of Egypt, its face 
tvould not be recognizable today. 

Frost action, even by itself, is an important factor in mountains and 
n temperate and cold climates, in helping wear away projecting rock 
nasses; the accomplishment of a single year may be small, but in long 
iges the accumulated results are of great magnitude. Further, it pro- 
vides smaller fragments upon which chemical attack may be made more 
effectively. But it must not be assumed that all heaps of broken " slide 
rock " are caused only by frost action, for several other processes of 
rock disintegration produce similar results. 

Changes in Temperature. Most substances when heated expand, 
md if they are good conductors of heat like metals, the expansion ex- 



tends well beyond the actual place of heat application. An iron rod 
cannot be held with the free end in the fire without discomfort to the 
hand, but a piece of rock can be so held. It is a poor conductor of heat. 
If the heated end expands and the rest does not, the two must separate. 
When cooling occurs an outside layer that contracts becomes too small 
for the unchanged interior; it cracks. A similar process goes on, but 
much more slowly, when rocks are heated by the sun and contract in 
the shade (Fig. 11). On upstanding rock masses the projecting corners 

Fig. 11. Buckling in sandstone layers due to expansion from heating by the sun. 
Wyoming. (U. S. Geol. Surv.) 

are spalled off first, then thin shells peel off or exfoliate, like onion layers, 
as illustrated in Fig. 12. Boulders become spherical and gradually dis- 
appear. Larger masses take on a rounded form, like some dome-shaped 
hills. The scales that drop off become further broken up and dust-like 
material results. Similar effects, produced in a much shorter time, are 
seen in a stone building that has passed through fire. Dr. Livingstone 
found in Africa that rock surfaces heated to 137 by day, upon cooling 
rapidly at night threw off with sharp reports angular fragments up to 
200 pounds in weight; and according to Stanley, contraction caused by 
cold rain falling on those sun-heated African rocks causes them to split 
and exfoliate. Still another though minor effect is produced by change 
of temperature; unlike mineral grains composing a rock neither absorb 
heat nor expand equally; consequently minute interior strains are set 
up that eventually disintegrate a surface layer of the rock. 

Thus disintegration proceeds, and broken rock results. Rain or wind 
may carry away the disintegrated material and leave fresh rock surfaces 



.Fig. 12. Exfoliation, of scaling of rock, by alternate expansion and contraction of 
surface layers. Nevada City, Cal. (U. S. Geol. Surv.) 


exposed to further attack. And so the process repeats itself until ex- 
posed rock surfaces finally disappear. Hills or cliffs are thus worn down, 
or jagged valley slopes are smoothed into gentle curves. 

The process is most rapid in regions where there is a pronounced differ- 
ence in temperature between day and night or between winter and sum- 
mer. In temperate climates the difference between the heat of sum- 
mer and the cold of winter may be 100 or even 150. And in desert 
climates the daily range may be 50 or exceptionally 100. Therefore 
high latitudes and altitudes or semiarid regions favor disintegration 
by temperature changes more than do regions with a moist, equable 

Rain. As compared to the agencies previously discussed the me- 
chanical work of rain is small, although it is an important agent of chem- 
ical weathering and of erosion. The impact of rain on steep slopes com- 
posed of soft rocks tends to dislodge small particles which are readily 
washed or blown away. New surfaces are thus exposed to a repetition 
of this process, which in arid regions helps to form " badland " topog- 
raphy. The mechanical work of rain is also seen in the softening of 
rocks or clays, enabling them to be more readily removed by rain wash 
or streams. 

Wind, when it causes sand blasts, is also a mechanical agent of abra- 
sion, but its work is of such broad scope that it is treated separately in 
later pages. 

Plants and Animals. The wedging apart of rocks by the roots and 
trunks of growing trees is a familiar sight to the observer of nature. 
In their slow expansion by growth they exert a powerful disruptive force 
which even large masses of rock are unable to withstand. Likewise, 
the rootlets of plants and shrubs insinuate themselves into little crevices 
of bedrock and boulders, and slowly break them apart. In every little 
crack where seeds may lodge and grow this process goes on. In the 
course of time the amount of work done in this way must be great, 
but its chief importance is in helping chemical attack to proceed more 

Animals such as moles, gophers, worms, and ants, by making holes 
and burrows in the soil, aid disintegration to a slight degree. Their 
work enables the other agencies to come more readily into contact with 
covered rock surfaces. The most destructive animal, however, is Man. 
He has felled the forests and otherwise destroyed the protective cover- 
ing of plant life over the soil, allowing it to be carried away by rain or 
wind and thus expose fresh surfaces of rock to processes of disinte- 



Chemical agencies, like mechanical agencies, are essentially super- 
ficial in their work. They act only on a thin skin of the outer rocky 
crust, and lose their effectiveness at shallow depths. Their work is 
inextricably interwoven with that of the mechanical agencies, and it is 
only for convenience of discussion that we consider them separately. 
The result of their combined attack is the soils from which we derive our 

Decomposition or decay may act directly upon solid bedrock, but it 
proceeds much more readily upon rocks previously fragmented by disin- 
tegration processes. It goes on more slowly than disintegration, and is 
favored by jxtojsture and heat-j-aad-retaFded-by cold.. Consequently, 
unlike disintegration, it is promoted by warm, moist climates, by mod- 
erate relief, and by a covering of vegetation. If the climate of Egypt 
were less dry, fewer pieces of old stone art would have been preserved 
for us; and if the climate of Labrador were less cold, a productive soil 
mantle might exist there. 

The chief agencies that bring about decomposition of the rocks are 
oxygen and carbon dioxide from the air, water, the organic acids re- 
leased by decaying vegetation, and certain bacteria. 

Decomposition. If a piece of old iron is left exposed in a damp cli- 
mate, it rusts; that is, it changes from iron, which is unstable in the 
presence of oxygen and water, and entering into combination with these 
substances, forms rust or limonite (iron, oxygen, and water), which is 
stable under the changed conditions. A simple chemical reaction has 
taken place, and a new substance has resulted. This is one of the changes 
that occur when rocks decay. The iron-bearing minerals of the rocks, 
for example, yield limonite, and the rock surface or resulting soil takes 
on the familiar tints of iron oxide, imparting to deeply weathered areas 
the red, yellow, or brown colors so pleasing to the eye. 

But to produce fertile soil, other rock minerals must be decomposed. 
The -chief rock-making minerals are the feldspars, and they are also im- 
portant soil-yielding minerals. The chemical change of one of them, 
orthoclase/may be illustrated: 

OrthoclaseH- Water + Carbon dioxide yields Clay -f- Silica + Potassium carbonate 
a = (Various hydrous aluminum silicates) -f-SiOa 

This is one of the most important reactions that take place in nature, 
since the existence of life so largely depends upon it. The necessary 
carbon dioxide is obtained from the atmosphere and in part from. decay- 
ing vegetation^ and enters readily into solution with water to form car- 


bonic acid- Clay is an essential ingredient of good soils, and potassium 
carbonate is a necessary food of plant life. The potash chemically locked 
up in orthoclase is thus set free as a soluble food that can be assimilated 
by plants. 

Similarly, other common rock-making minerals that contain alumi- 
num, such as amphibole, chlorite, and mica, yield clay. Some also 
yield potash. Quartz, being relatively insoluble, resists decomposition 
and remains in the soft decomposed materials as grains of quartz, or 
sand. When granite, composed chiefly of feldspar, and quartz, is 
acted upon, the resulting soil is a mixture of clay and sand grains, 
called loam. 

The combination of a substance with oxygen is oxidation; with 
water, hydration; and with carbon dioxide, carbonatization; all these 
processes enter into decomposition. As a result new substances called 
oxides, hydrates, or carbonates, are formed. Some remain behind in 
the soil, others may be carried off in solution. Hydration is usually 
accompanied by an increase in volume, and this swelling is an additional 
factor in helping to break up the rocks. Some geologists think that this 
effect of hydration is very important in producing some of the effects 
of disintegration ascribed to change in temperature. 

Solution. In addition to the more complex chemical changes men- 
tioned above, simple solution also aids chemical decomposition. Pure 
water (H 2 0) is a poor solvent for rock minerals, but when it is combined 
with carbon dioxide (C0 2 ) to form carbonic acid (H 2 C0 3 ) it is a powerful 
natural solvent. It attacks, among other substances, calcium carbonate 
(CaC0 3 ) and converts it into calcium bicarbonate [H 2 Ca(COs)2], which 
is quite soluble in water. Some rocks, such as limestone, are composed 
almost entirely of calcium carbonate, which forms the mineral calcite, 
and in others, such as sandstones, this substance acts as a cement to bind 
the individual sand grains together. When carbonic acid acts on such 
a rock, the binder is dissolved, the grains loosen, and the rock crumbles 
and breaks down into sandy soil. 

In the case of limestone, the greater part of the rock may be slowly 
dissolved, leaving behind only the insoluble impurities, usually clay, 
to form a residual soil. Unbelievable thicknesses of rock have been 
removed by this process, at a rate that in places may be as great as 1 inch 
in 25 years. The impurities gradually accumulate on the surface, giving 
rise to the fertile soils of such limestone regions as southern Kentucky. 

The impurities in the limestone may, in places, consist of iron or 
manganese compounds instead of clay, and residual accumulations of 
these substances have formed valuable deposits of iron or manganese 
ore that are being mined in different parts of the world. 


Solution is also effective in removing some of the soluble products 
that result from chemical decay. 

Plants and Animals. Both plants and animals aid in the chemical 
attack upon rocks. Plants in their growth extract-carbo-nr-diaxidejrom 
the atmosphere; they release the oxygen and retain the carbon. On 
their death and decay some carbon dioxide is "released and furnishes a 
source of carbonic acid for cHemical attack- upon 'the~unHeriying rocks. 
As the plants dfe^andTothers grow upon their ruins, a part of the carbon 
may accumulate, as a carbonaceous residue, known as humus, desirable 
as an ingredient of fertile soils. 

The humus also supplies to the rain water that seeps through it small 
quantities of complex organic acids which are in themselves effective 
chemical agents of rock destruction. They also supplement the work of 
carbonic acid in mineral decomposition or solution. 

In addition, these acids are potent reducing agents; that is, they are 
able to take away oxygen from some oxidized compounds. This is 
most strikingly seen in the effect upon the ferric hydroxides that color 
the soils red, brown, or yellow. The ferric hydroxides (Fe 2 3 + water) 
are reduced to ferrous oxide (FeO), some of which unites with carbon 
dioxide to make ferrous carbonate, and this in turn is soluble in carbonic 
acid. Thus some of the iron is leached out and removed, and the re- 
maining part no longer gives high colors to the rocks or soils. This is 
strikingly seen where red sandstones overlain by vegetable mold are 
grayish in hue just beneath; and where rootlets extend down into red 
rocks there is a bleached zone surrounding them. Where ferric oxides, 
limonite or hematite, form the cement that binds the grains of sedimen- 
tary rocks together, its removal by the action of organic acids causes 
the rock to fall to pieces and so brings about the formation of soil. 
Where much organic matter is present in soils, vivid colors are absent. 
The two are incompatible. For example, in certain clays dark blue, 
dark gray, or greenish to black color denotes the presence of organic 
matter; and any iron present must be in the colorless ferrous form. 
But if these same clays are fired to make bricks, the organic matter is 
burned out, the ferrous iron is oxidized to the ferric state, and red brick 
results. If iron is not present, white brick is formed. 

Insects and other animals that live and move about in the soil aid 
chemical work by upturning the soil and exposing fresh surfaces to weath- 
ering. Their openings in the soil also enable weathering agents to reach 
more readily the underlying bedrock. Thus, Darwin states that in 
England earthworms bring to the surface 10 tons of mold to the acre 
every year, and Branner believes that in many tropical regions ants are 
even more effective in upturning the soil. 


Bacteria also contribute their part. Certain forms have the unusual 
power of assimilating carbonate of ammonium and setting free nitric 
acid which attacks and decomposes rock minerals. They penetrate in 
great numbers every little nook and cranny in the soil or bare rock, and 
in time their effect is of no inconsiderable geological significance. 

The general effect of chemical decomposition^^crbreak up the com- 
plex rock minerals to form softer incoherent mSerials that may remain 
as a soil mantle, or may readily be carried away J)y_ rain or wind. 


The results of weathering are no less numerous than are the physical 
and chemical agencies that bring them about. It is unusual, as we have 
seen, for one alone of these agencies to prevail for any time; they operate 
together. The sum total of their work destroys the integrity of the sur- 
face rocks, fashions the smaller details of the landscape, and spreads a 
life-giving soil mantle over most parts of the earth. By means of them 
some of the rugged details of mountain scenery have been sculptured 
and the gentle slopes of softer vistas have been smoothed. Materials 
are supplied to build flood plains and deltas and sedimentary rocks. 
Some of these results of the cooperative work of the agencies of weath- 
ering may now be considered, and of these, soils stand first. 

Soil and Rock Mantle. Nearly everywhere the Earth's surface is 
mantled by a thin veneer ,ol soil or of disintegrated rock. Its thickness 
' varies considerably; compared with the Earth as a whole, it is but a film. 
In tropical regions it may extend a few hundred feet in depth. Nor- 
mally it is only a few feet, or a few tens of feet in thickness. Soil is 
formed, as we have seen, chiefly from the breaking up of rocks; de- 
cayed vegetation in most places adds to it but little. In the agricultural 
sense the term soil is applied only to that upper portion or topsoil 
which contains some humus and supports plant life, but in the geologic 
sense it embraces as well the underlying rotted rock or subsoil. 

Soils may be divided into two large groups: residual soils, which 
have been formed in place from the immediately underlying rock; 
and transported soils, which have been moved from their place of origin. 

Residual soils normally pass gradually downward from a topsoil 
supporting vegetation into the subsoil of coarser material full of bits of 
rotted rock, and then imperceptibly into decayed rock that crumbles 
more or less easily, and this in turn merges into unaltered solid bedrock 
(Fig, 13). This gradual transition from topsoil above to solid rock below 
is one proof that the soil has been formed in place by the decomposition 
of the underlying rock. . The transition between rock and soil is not 



Fig. 13. -*- Residual soil formed by rock weathering and decay. The material graduates 
from firm rock below, through rotten rock and then subsoil, to true soil above. The 
transition is gradual. The soil above is colored dark by decayed organic matter (Mer- 
rill, U. S. Nat. Mus.) 


only gradual but also highly irregular from place to place. Rich re- 
sidual soils are widespread over most of the Piedmont district of the 
southeastern United States, and are the chief soils in humid tropical or 
subtropical regions. 

Residual soils in general are composed of much finer materials than 
transported soils, for in them decay has been more complete. 

Transported soils have been shifted by rain or running water (al- 
luvial), by wind (eoliari), by moving ice sheets (glacial), or by gravity 
down hill slopes (colluvial). Consequently they vary in composition 
more than residual soils and range from the finest silts to gravel. They 
have no essential kinship with the underlying rock and their source may 
have been far distant from their present site. Wide areas of the north- 
eastern and north central United States are mantled by glacial soils that 
came from farther north, and the fertile silts that flank the lower Mis- 
sissippi River came from far upstream. 

Character of Soils. The character of soils, whether they are residual 
or transported, depends largely upon the kind of rock from which they 
were derived and upon the climatic conditions under which weathering 
took place. Thus rocks composed chiefly of feldspar under arid con- 
ditions disintegrate to feldspathic sand; under humid conditions they 
yield clay. A pure sandstone yields only sand. According to the size 
of the particles which compose the rock mantle, the following gradations 
are recognized : Pieces of loose rock from the size of a small melon up are 
termed boulders; those larger than peas are called gravel; pieces smaller 
than peas, but not coherent when wet, are sand; and the finest material, 
which can be carried by the wind, is dust; the last is termed silt or day, 
according to its character, and generally coheres when wet. Ordinary- 
soils are composed of variable mixtures of sand and these finer materials. 

Loam, a mixture of clay and sand, is easily worked and makes ex- 
cellent soil; pure clay contains abundant plant food but is apt to be stiff 
and difficult to work; very sandy soils are usually unfertile. 

Soils very rich in humus are called muck. When a clay soil contains 
a considerable quantity of calcium carbonate it is termed a marl. Thus 
sands, loams, clays, mucks, and marls are the chief kinds of soils and there 
are all gradations of these into one another. Owing to the presence of 
the dark organic matter or to the greater oxidation of the iron compounds, 
and often for other reasons, the topsoil is likely to be much more strongly 
colored than the underlying subsoil. 

Talus. On mountains, where the prying action of frost, or ex- 
foliation by changes of temperature, or swelling by hydration take 
place, the rock so broken collects by gravity on the lower slopes, form- 
ing slide rock or talus. This is illustrated in Figs. 10, 14. The talus 



may be a sheet-like form that flanks the mountain slope, as in Fig. 10, 
or it may be cone-shaped, when it is called a talus cone, as in Fig. 14. 
A mountain may become so buried in its own talus that only its summit 
projects. Those who have climbed mountains realize the abundance 
and size of talus slopes, particularly in high latitudes. They are, how- 

Fig. 14. Mt. Washington, Oregon, showing talus cones. (U. S. Army Air Corps.) 

ever, but transient features, for the materials composing them eventu- 
ally are comminuted and are largely carried away by water or by wind. 
Residual Boulders. The change of bedrock into soil is apt to pro- 
ceed first along cracks and fissures. These usually intersect each other 
and inclose large and small blocks of bedrock, Consequently the agents 
of decomposition and disintegration begin their work on the outside of 
such blocks. The vulnerable corners succumb, then the sides are at- 
tacked; rounded residual boulders may result. Where the attack of 
chemical and mechanical agencies has gone so far that rock masses or 
boulders take the shape of spheroids, it is called spheroidal weathering. 


Concentric layers, like the skins of an onion, may continue to peel off 
until eventually the rock bodies disappear. Also, masses distributed 
through the bedrock may be different in composition or texture from 
the rest, and thus harder or less soluble. These may also be left as 
residual boulders (Fig, 1 5) , Some residual masses may rest in apparently 

Fig. 15. Residual boulders left by decomposition and wearing away of bedrock. The 
boulders are included masses of a harder, more resistant material and of rounded shapes 
(concretions). This shows that some residual boulders differ from the bedrock on which 
they lie'. Coalinga, Cal. (U. S. Geol. Surv.) 

unstable positions (Fig. 16). It must not be assumed, however, that all 
boulders are formed by this process, for many are transported blocks 
whose composition is unlike the underlying rock, and many have been 
formed by differential disintegration or wind erosion. 

Exfoliated Forms. Exfoliation has also operated on a grand scale, 
not to be compared with the thinner veneers of disintegration by sphe- 
roidal weathering, to produce great picturesque domes such as those of 
the Yosemite Park and Stone Mountain, Georgia (Fig. 303). On Stone 
Mountain huge slabs, scores of feet across and up to a foot in thick- 
ness, are in various stages of peeling off; doubtless hundreds of feet 
have in the past been removed from the top and sides of the mountain 
in this manner. Chemical decomposition seems to have played little 



part in producing this result, since the loosened material, though fri- 
able, is quite fresh. Changes in temperature must have been the domi- 
nant agent of destruction, 
though frost action and hydra- 
tion may have played a minor 

Uifferential Weathering. 
Nature has tooled rocks of dif- 
ferent resistance to erosion into 
grotesque forms and scenic bits 
of landscape such, for example, 
as may be seen along the Cody 
entrance to Yellowstone Park. 
Softer or more soluble parts 
have been removed and the 
harder resistant parts are left to 
stand out in relief (Fig. 107). 
They may be only small ribs of 
insoluble quartz that project 
from the surface of soluble lime- 
stone, or " mushroom rocks " or 

pagoda-shaped boulders. Or they may be larger features such as bal- 
anced rocks (Fig. 16) and pinnacles or columns, or even a succession of 
cliffs and benches where hard and soft layers of rock alternate, such as 
may be seen on a grand scale in the Grand Canyon of Arizona. The 
more common balanced rocks, pinnacles, and pedestals, such as attract 
the visitor in the Garden of the Gods or Monument Park, in Colorado, 
are formed usually by the more rapid weathering, chiefly by mechanical 
agencies, of softer beds of rock that underlie harder strata. The caps 
of harder rock eventually fall to the ground and, until they are disin- 
tegrated, remain as boulders. 

Fig. 16. Residual boulder resting on 
bedrock. "Balanced Rock," Garden of the 
Gods, Colorado. 


The geologic work performed directly by the wind is mechanical in 
its nature; it is a builder as well as a destroyer. It carries away prod- 
ucts of weathering; it sweeps sand against rocks and abrades them; 
and it builds up dunes and dust deposits. Its work is most effective 
where vegetation is absent; consequently its results are to be seen in 
arid or semiarid lands and near sandy shores. The effect of the wind 
in forming waves that waste the coasts may best be deferred to the dis- 
cussion of oceans. 


The Wind by Itself. The wind by Itself plays a minor role in blow- 
ing away the products of weathering as they are formed, thus exposing 
fresh rock surfaces to further wasting. With its aid, therefore, disin- 
tegration and decomposition can proceed faster. This action proceeds 
best on hillsides or valley slopes where lodgment of weathered particles 
is insecure and where gravity aids; but it also operates even in flattish 
areas, particularly in arid regions, where mechanical agencies disintegrate 
the surface, and wind blows the material away, making shallow T depres- 
sions. Such wind work is termed deflation. In semiarid regions such as 
Nebraska, in times of drought, the wind dries the soil and blows it 
away from cultivated fields, with disastrous results to young crops and 
the supply of soil. 

The wind also transports vast quantities of material. A gentle breeze 
lifts and carries dust, a strong wind drives sand (Fig. 17), and a tempest 

Fig. 17. Sand storm sweeping over Khartoum North; in front is the Blue Nile. 
Shows enormous transporting power of the wind. Soudan, June 6, 1906. (Win. Beam.) 

can move gravel the size of peas. For example, a single storm that 
travelled from the arid Southwest a thousand miles to the Great Lakes 
region brought with it a million tons of dust, which was deposited on 
the snow over a wide area. Such material, dropped when the wind 
slackens, may eventually form deposits of great magnitude. They are 
known as eolian (Aeolus, god of the winds) deposits, to distinguish 
them from water-laid deposits. 

Abrasion by Wind-blown Sand. The smoothing effect of sandpaper 
applied to a rough surface is a familiar example of abrasion. The 



particles of sand fastened to the paper make sharp and effective cutting 
tools. Similarly, begrimed stone or brick buildings today are being 
renovated by a sandblast sand blown by compressed air. Nature 
supplies a similar agent of abrasion in the sand moved by wind. Though 
less efficient as a cutting tool than its artificial counterpart, it is, over long 
periods of time, an important agent of rock destruction. 

The rate of cutting by storm-driven sand may be quite rapid. For 
example, window panes of houses along seashores have lost their trans- 
parency in a day or two, and have been completely penetrated in a 

Fig. 18. Rounded rock forms produced in part by wind erosion. Note irregularities 
in surface. Arizona. (Bateman.) 

month or so. Wooden telegraph poles planted in the desert are in some 
localities quickly cut down by the sand drifting past their bases, and on 
stretches of railroad that cross deserts, rails have to be replaced fre- 
quently because of abrasion by drifting sand. 

The most effective abrasion by wind-driven sand is to be seen in arid 
regions and deserts, where rainfall is scanty and protective vegetation is 
sparse. Exposed rock masses are carved and sculptured into rounded 
(Fig. 18) and " hoodoo " forms. In detail, the surfaces are commonly 
pock-marked or intricately fretted by the removal of softer parts. Fur- 
ther evidence of eolian abrasion is to be seen in the smoothed, polished 
pebbles that lie in such areas. Many of them are faceted into angular 
forms, commonly with triangular cross sections. Such pebbles are called 



Sand Dunes. Sand dunes are hillocks made by wind-borne sand in 
a manner similar to that in which snow forms drifts. Their growth is 
started by some obstacle, such as a rock, a bush, or a surface irregularity, 
that gives local protection from the wind or causes wind eddies, so that 
the moving sand is deposited. A dune, once started, continues to grow 
and may reach a height of 100 feet to 200 feet or more. Even the 
erection of buildings starts their formation in some places. The surface 
of a dune is commonly covered with fine parallel ridges of sand an inch 
or so in height, transverse to the direction of the prevailing wind, and 
called ripple marks because they resemble the markings on sand made 
by water ripples (Fig. 19). Dunes are formed wherever free-moving 

Fig. 19. Sand dunes near Mammoth Station, CaL, showing ripple marks, 

Geol. Surv.) 

(U. S. 

sand exists and where there are prevailing winds. Frequent changes in 
wind, no less than vegetation or exceptionally rough topography, are 
unfavorable to their development. They are found in all parts of the 
world along low coasts where sand made by waves is washed ashore and, 
caught up by the prevailing winds, is drifted inland and accumulated. 
Dunes so formed occur at various places along the coasts of the United 
States; in England; on the shores of the Baltic Sea; in Holland, Bel- 
gium, and France. Similarly they are produced on the shores of large 
lakes or inland seas; the southern end of Lake Michigan is fringed with 
high sand dunes. They are also of common occurrence along broad 
river bottom lands in semiarid regions, as along the Arkansas River in 
western Kansas. Sand dunes are especially abundant in arid regions, 
where disintegration is rapid and a protective mantle of vegetation is 



lacking. Large areas in the great deserts of Africa, central Asia, Arabia, 
Australia, and western America are covered by them. It is a mistake, 
however, to suppose that all desert lands are mantled deeply with sand. 
Even in the Sahara there are great areas of barren rock from which the 
wind sweeps all loose material. 

Shape of Dunes. The shapes of dunes vary according to the in- 
tensity and direction of the wind. Some are linear, and others very 
irregular; many are curved or crescent-shaped in plan and are called 
barchanes. The windward side of a dune has a gentle slope which is 
determined by the strength of the wind. The leeward side, capped by 
a rather sharp crest, has a steeper slope, which is the angle of repose of 
free falling sand, and varies from about 20 to 30. The shape thus tells 
the direction of the prevailing winds. 

Migration of Dunes. A dune once formed is never stationary unless 
it meets an object larger than itself, or is held by a mat of vegetation. 

Church of Kunzen 

In 1809 

Place of buried church 

"Ruins of Church 

Fig. 20. Movement of a sand dune during 60 years on the east shore of the Baltic 
Sea at the village of Kunzen. (After Berendt.) 

It moves ceaselessly onward with the prevailing winds, by the slow trans- 
ference of sand grains from the windward to the leeward side. During 
this progress its height is maintained or increased, though its outline 
may suffer change. Thus, along coasts where the prevailing winds are 
from the sea, dunes may form a belt extending several miles inland. 
Their rate of movement may average about 20 feet a year, as in Den- 
mark, or reach more than 100 feet a year, as on the Biscayan coast of 
France. In their march they cover and destroy arable lands, forests, 
and even towns, leaving desolate sandy wastes behind (Figs. 20, 21). 
In the deserts of central Asia, Sven Hedin, the explorer, found ruined 
cities of an ancient civilization emerging from sand dunes which for a 
long time had overwhelmed them and the fertile lands that once sup- 
ported them. 



The devastating migration of dunes has been mastered along the 
coasts of European countries by the skillful planting of trees or other 
vegetation to prevent the removal of the sand particles from the wind- 
ward slope. The dunes then become stationary. In France the plant- 
ing of forests on the dunes resulted not only in preventing dune migra- 
tion but in the raising of profitable forests on otherwise waste land. 
In arid regions, however, where vegetation will not grow, the migration 

Fig. 21. Forest overwhelmed, killed, and then left exposed by marching sand dunes. 
Manitou Island, Lake Michigan. (U. S. Geol. Stirv.) 

of dunes is almost impossible to control, and nothing should be built in 
the path of their travel. 

Loess. Extensive deposits of fine clay-like materials have been 
built up chiefly by the wind to form loess. The name comes from Alsace, 
where it designates a peculiar fine-grained, yellowish-brown, lightly 
compacted earth. It is remarkably fertile. The deposits show no 
horizontal banding, but form upright bluffs because of a tendency to 
cleave vertically. These deposits are extensive along the Rhine, oc- 
curring well up on the mountain slopes, and form the most fertile soils 
of central Europe. They cover tens of thousands of square miles in the 
central Mississippi Valley region, especially in Iowa, Nebraska, and 
Kansas, and are found also in Oregon and Washington. Similar ex- 
tensive loess deposits occur in the rich pampas of the Argentine. They 


are thick deposits that have formed chiefly from accumulations of dust, 
which in America and Europe was probably supplied by the flood plains 
of glacial streams during the recent ice age. There was no vegetation 
to hold the soil at that time, and the wind blew away the finer particles. 
The greatest development of loess is in Asia, particularly north central 
China, where it covers 230,000 square miles, and the yellow earth washed 
down from it has given the Yellow River and the Yellow Sea their names. 
It is believed by Von Richthofen to have been carried by the wind from 
the great deserts of the interior, especially the Gobi desert, and to have 
spread over the adjacent area. Its thickness reaches 300 feet or more. 
It is remarkably fertile, and even though it has been cultivated for thou- 
sands of years, it is still productive. Streams have cut canyons into it, 
and at the base of the bluffs the humbler Chinese have fashioned in it 
cave dwellings which they have used for centuries. Where roads cross 
it, the material loosened by the cart wheels has been blown away, and 
this going on for centuries has caused the roadways to sink lower and 
lower, so that some of them are now small canyons. 


1. A Shorter Physical Geography; by Emmanuel de Martonne, translated by 
E. D. Laborde. 338 pages. Alfred A. Knopf, New York, 1927. 

Excellent discussion of the elements of climate, pleasingly written. 

2. Introductory Meteorology; issued by the National Research Council. 150 
pages. Yale Univ. Press, New Haven, Conn., 1918. 

The principles of meteorology. 

3. College Physiography; by R. S. Tarr, edited by Lawrence Martin. 837 pages. 
Macmillan, New York, 1914. 

Part III is a good section on the work of the atmosphere. 

4. Rocks, Rock Weathering, and Soils; by G. P. Merrill. 400 pages. Macmillan, 
New York, 1913. 

Well written, authoritative treatise on the breaking up of rocks and soil formation. 


The Rainfall. Probably there are no parts of the Earth's surface on 
which rain does not sometimes fall. Death Valley, in California, one 
of the driest places in North America, has about 2 inches of rainfall 
annually. Even the most arid spots in the Libyan Desert have had 
rain at least once within a measured span of 12 years. But the amount 
of rainfall a country receives is dependent on a variety of factors, such 
as the direction of the prevailing winds, the nature of the places over 
which they have previously passed, and the height above the sea of the 
country which receives them. Thus it happens that the amount of 
rainfall received by the land is very unequally distributed over the 
world; in many places, as in Central America, it may be as much as 
100 inches per year and in parts of India 500, but in the great deserts 
it is less than 10. In North America, it is true in general that in the 
Atlantic seaboard region and in the southern states the rainfall is 40 
inches or more per year; westward to the Mississippi River it diminishes 
to 30 incnes or somewhat more; in the Great Plains region to 20 or less; 
in the Basin and Range country between the Rocky Mountains and the 
Sierra Nevada to 10 or less. Locally in the mountains it is increased, 
because mountains are great condensers of moisture. On the Pacific 
coast it increases again. Roughly speaking one may term as arid those 
regions where the rainfall is less than 10 inches, as semiarid those where 
it is 10 to 20 inches, and as humid those where it is more than 20 inches. 

Various things happen to the rainwater after it reaches the surface 
of' the land. A part is trapped by surface depressions and remains in 
lakes and ponds. A part evaporates, a part sinks below the surface 
and becomes ground water, and a part flows down the slope of the sur- 
face, forming the immediate run-off. 

The Run-off. The proportion of the rainfall that contributes to 
the run-off in a given region depends on several factors such as (1) sur- 
face slope, (2) porosity and solubility of surface material, (3) character 
and amount of vegetation, (4) temperature and humidity of the atmos- 
phere, and (5) distribution of the rainfall throughout the year. It 
therefore varies greatly from place to place. Certain areas of very- 
soluble and porous limestones have essentially no run-off, all of the rain- 



fall sinking below the surface, whereas on steep slopes in the Appalachian 
Mountains, nearly all of the water from rains in the early spring flows 
off over the surface, because the ground is already saturated with melted 
snow. On an average, about one-fifth of all the rain that falls forms 
run-off. Since much of this fifth flows from areas thousands of feet 
high to the sea, the power of the resulting streams to do work is very 
great. As the water under the influence of gravity flows off over the 
surface during and after rains, it seeks the shortest and steepest route 
downward, following the lowest depressions wherever they present 
themselves, and is continuously augmented by seeps and springs repre- 
senting the reappearance of the ground water which sinks below the 
surface at higher levels. 

Stream Valleys; Stream Erosion. Every observer, as he stands 
on the brink of the Grand Canyon of the Colorado and looks down at 
the winding ribbon of the river a mile below, is impressed with the vast 
size of the rock-walled valley and with the puny appearance of the stream 
at its bottom. The earliest observers thought that the Colorado had 
found this low course already prepared for it by a great gash or rift in 
the Earth's crust. But the visitor of today, if he knows some of the 
principles of geology and understands rivers and their behavior, can be- 
lieve that the river once flowed through a shallow trough at the level of 
the canyon's brink, and that very gradually the sediment-laden water 
cut down little by little into the underlying rocks, sawing ever more 
deeply until in the course of time it dug a trench a mile deep. This same 
visitor surveys from an eminence the intricate pattern of 'the canyons 
and sub-canyons which empty into the main valley, arranged symmet- 
rically like the veins in a leaf, most of them entering flush with the main 
stream. He realizes that they are not fortuitous, but are all parts of a 
unified and interrelated system developed in obedience to a common 
law, and even having a common destiny. And when he sees this same 
pattern repeated again and again in stream systems large and small 
throughout the world, he concludes that all rivers act according to certain 
definite rules imposed on them by their surroundings. The fundamental 
cause of stream flow is the pull of gravity upon mobile liquid at the 
Earth's surface. But as water flows it picks up loose rock particles, 
carries them downstream, and by their aid wears away the solid rock of 
its bed. This process of wearing away is called erosion, whether the 
active force be a stream, a glacier, the wind, or some "other agent. 

Development of Streams. It is probably true that no part, of any 
continent has escaped inundation by the sea at one time or another 
during the Earth's history. It follows that not infrequently areas of 
shallow sea bottom (such as the bottom of Hudson Bay) have been 


warped up above sea level and have become dry land. As soon as such 
an area appeared above water, it would have received rainfall, and the 
run-off would have been shed from the high places toward the low, and 
so into the sea. All the conditions necessary for stream erosion would 
be fulfilled a quantity of water being pulled downward by gravity, 
with loose mantle rock constantly being formed by weathering processes, 
ready to be picked up, carried away, and deposited at lower levels by 
running water. 

Consequent Streams. Consider the history of a land area newly 
emerged from the sea. During the first rainfall following emergence, 
the impact of raindrops upon the surface loosens the exposed fine par- 
ticles of mantle rock and the run-off carries them down the nearest slopes 
in little temporary rills or rivulets. Where rivulets join at the bases of 
converging slopes the wash of their combined volumes picks up more 
loose particles and digs out gullies. Adjacent gullies join at lower 
levels and the constantly increasing volume of water excavates ravines. 
These in turn join to form even larger valleys. In this way, with many 
more junctions and convergences, the run-off reaches the sea in the form 
of streams loaded with sediment picked up at higher levels. The 
slopes down which the run-off flows in this early stage are called initial 
slopes, and the streams following or " consequent upon " these slopes 
are known as consequent streams. The rate of flow of the consequent 
streams depends entirely upon the steepness of the initial slopes which 
in the present case are gentle because they represent an upraised sea 
floor. Steep or gentle, however, slopes are always present; and once 
they have instituted a stream system, the streams will enlarge their 
valleys and develop new tributaries, actively eroding the land mass until 
they have brought the entire land surface nearly to the level of the sea. 

Since erosion is the wearing away of rock material, erosion by streams 
therefore involves (1) mechanical wear by rock fragments carried by 
flowing water; (2) solution of the rocks in stream beds; (3) picking up 
of worn particles; and (4) transportation of the mechanically worn and 
dissolved substances. Since weathering is practically universal the 
' comminuting action of the weathering process greatly aids erosion by 
providing streams with an abundance of loose material for ready trans- 
portation. However, a certain amount of deposition of transported 
material is always going on even in the smallest gullies where erosion is 
rapid, and thus deposition is a universal companion process to erosion. 

The rivers then are the main channels of drainage, and they are the 
chief factors in carrying away the waste of the land. They are the great 
trunk lines of transportation for the rock debris delivered to them by 
their tributaries. In addition they are themselves powerfully capable 


of wearing away the land through which they flow; in them the work of 
running water as a geological agency is most conspicuously displayed. 
If we should think of a typical river we should imagine it rising in lofty 
mountains through the union of many impetuous streams or dashing 
torrents; gathering headway it rolls rapidly through the belt of lower 
hilly country and emerges upon wide plains through which it wanders 
in many curves in a quiet and steady flow to the sea. The slope of its 
bed, its gradient, which may be as much as 20 to 30 at the head, becomes 
less and less until it is nearly horizontal at the river's mouth. 

Although we think of this as the ideal course of a river (and it is 
typical of many of the great rivers of the world such as the Amazon 

and the Ganges, and of a great num- 
ber of smaller ones such as the Po 
in Italy), we constantly find varia- 
tions from this type. Thus the 
Fig. 22. - Profile of a normal river show- Mississippi does not rise in a moun- 

ing its gradient. , -. , i , i 

tainous country, but in a moderately 

elevated region of low relief, and it has a very low gradient to the sea. 
On the other hand, some rivers rise in mountains near the sea, and hence 
their lower plains districts, corresponding to b of Fig. 22, may be short or 
wanting. In North America, the rivers of the northern Atlantic coast 
are mostly between these extremes and their courses lie between a 
and 6; those of the southeastern states are more nearly typical, since 
they rise in the Appalachian Mountains and flow out upon the Atlantic 
coastal plain. 


Associated with the destructive or degradational process of erosion is a 
constructive or aggradational process of deposition of the eroded material. 
Thus the normal stream gradient (Fig. 22) is a concave curve partly 
because erosion is most active near the headwaters of a stream while 
deposition takes place chiefly in the lower reaches. Strictly speaking, 
erosion by streams consists of three distinct processes, (1) corrasion, 
(2) solution, and (3) transportation. The static process of weathering 
(Chapter III) usually precedes erosion, preparing rock waste for easy 
seizure by running water, and deposition always follows it. But these 
two processes are entirely distinct from erosion proper. 


Stream erosion is limited in its operation, being confined to the bot- 
toms and sides of the channels through which the water passes. A 
river may be compared to a sinuous, flexible, and endless file, ever moving 



forward in one direction, and by means of the moving sand or gravel 
rasping away the rock beneath and beside it, thus cutting an ever-deep- 
ening trench. This particular phase of a river's work is called cor- 
rosion. The effectiveness with which a river corrades depends on 
several closely related factors; (1) on the abundance and character of 
the tools with which the river has to work, (2) on the velocity of its 
current, and (3) on the nature of the bedrock with which it has to 

(a) The Rivers Tools. Clear water moving over rock surfaces 
has little erosive effect. It has a certain solvent power and may thus 

Fig. 23. River bed full of more or less rounded boulders, showing the tools with 
which the stream works. Big Creek, Haywood Co., N. C. (U. S. GeoL Surv.) 

slowly dissolve and disintegrate rocks, but in order to corrade, a river 
must have tools. These are supplied by the sand and silt which it 
carries, and by the gravel and pebbles it can move if swift enough, .either 
in its normal flow, or in times of flood (Fig. 23) . This material is supplied 
to the river chiefly by rain wash and by its tributaries, but the stream 
also obtains it directly by wearing and undermining the sides of its 
channel. If its banks are steep, or cliff-like, the debris which accumu- 
lates at the foot of such slopes is seized by the river and used as its tools. 
It is by the striking, bumping, and grinding action of this material, 


carried along by the current, that the river is able to cut away the rocks 
over which it runs and thus to deepen its channel. 

In this process the material carried by the river is itself necessarily 
worn, has its sharp angles removed and becomes rounded or spheroidal, 
a form characteristic of the river's tools. Thus, if we find river 
gravel which consists of hard, well-rounded pebbles, we infer the material 
has been transported a long distance; on the other hand if the gravel 
is composed of angular bits of rock, and its situation shows that it has 
been transported, we infer that the distance must have been short. 

Up to a certain point an increase in the amount of grinding material 
supplied to a river with a given velocity of current aids in its corrasive 
power. Beyond this point an increase is not effective for the reason 
that the strength of the current is so consumed in the operation of 
transporting that the check, given by a tendency to corrade, would 
cause the river to deposit instead of carrying material farther. Since 
corrasive power depends on the strength of the blows struck by the 
moving particles, it is clear that this in turn depends upon the mo- 
mentum, that is, upon their mass multiplied by the velocity. Hence 
for a constant velocity, the greater the mass of the particles that is, 
the larger and heavier they may be the more effective agents of erosion 
they become. Thus in a stream carrying intermingled grains of sand 
and dust-like particles of clay the sand is the most effective agent. 

(6) Velocity and Corrasive Power. It is obvious from the preceding 
paragraph that other things being equal, the swifter a current is the 
more rapidly it will corrade. For, in a given time, not only will the 
number of corrading particles passing over a rock surface be increased 
with a swifter current, but the fact that each particle is moving more 
rapidly will add to its effectiveness. Further, since a swift stream can 
carry larger particles than a slower one, and since these larger particles, 
owing to their momentum, strike the channel with greater force, it 
appears that a moderate increase in the velocity of a stream will result 
in a great increase in that stream's cutting power. Calculation of these 
complex factors shows that doubling a stream's velocity increases its 
corrasive power at least four times, and perhaps in some cases as much 
as sixty-four times. In other words, corrasive power varies by a factor 
between the square and the sixth power of the velocity. 

(c) Character of the Bedrock. Such rivers as the Platte and the 
Missouri on the Great Plains, the James, the Roanoke, and the Savannah 
on the Atlantic Coastal Plain, the Yangtse on the coastal plain of China, 
the Thames in the London Basin, the Seine in the Paris Basin, and a 
host of others besides all are turbid with sediment. They are alike 
in that they flow through regions of shales, clays, and soft sandstones. 


On the other hand most of the streams of New England and the Adiron- 
dacks, and of the Highlands of Scotland, together with such streams as 
the upper Danube, are relatively clear. And these are alike in that 
they drain regions of hard igneous and metamorphic rocks. Thus we 
find countless illustrations of the fact that the composition of the rock 
in a stream valley has a great influence on the rate of erosion. The 
structure of the rock masses also exercises a profound effect. If they 
are jointed and thinly bedded, many planes of weakness are present along 
which the stream can chip out and dislodge fragments. Vertical beds 
of shale present ideal conditions for this kind of erosion. If the rocks 
are massive, like granite, streams traversing them erode with difficulty. 


Even without visible tools running water wears aw T ay the land by 
solution. Its solvent action is greatly aided by substances already in 
solution, inherited in part from the ground water that issues upstream 
in springs and seeps and in part directly from the rainwater w r hich com- 
monly contains dissolved gases acquired from the atmosphere and from 
decaying vegetation. The rock most susceptible to the solvent action 
of stream water is limestone (calcium carbonate). But even where a 
stream flows over relatively insoluble rocks, certain soluble minerals are 
decomposed, thus loosening the adjacent insoluble mineral grains and 
preparing them for mechanical seizure by the current. 


The material carried by a stream forms its load. While the greater 
part of this is carried (a) mechanically in suspension, a very considerable 
portion is transported (6) chemically in solution, and still another part 
is (c) rolled or moved along the bottom. The ultimate goal of the river 
is the sea into which, at the end of its journey, its remaining load is 
deposited. The various aspects of this work demand consideration. 

(a) Material in Suspension. The size of the particles that a river 
is able to carry in suspension depends on (1) the character of a river's 
current, (2) its velocity, and (3) the relative weight or specific gravity 
of the particles. 

1. Character of the Current. If the mass of water forming the 
current moved forward in a perfectly uniform manner, each particle of 
water from side to side and from top to bottom moving forward with 
the same velocity as every other particle, only the very finest material, 
such as microscopic granules of clay, would remain any length of time 
in suspension. A sand grain dropped into the stream would sink to the 


bottom and there remain at rest, unless the stream were strong enough 
to roll it along. But the current of streams is not of this character. 
The more central portions are moving more swiftly, sliding over those 
toward the bottom and sides, while there is a constant interweaving of 
swifter sub-currents up and down and toward the sides and even back- 
ward, forming eddies or whirling movements. The whole effect is like 
the stirring of water in a glass. Sand at the bottom is quickly lifted and 
kept in suspension by these movements while it is carried along by the 
main current. When particles in suspension in water attain a certain 
degree of fineness their settling, even when the water is still, becomes 
very slow. Thus water such as that of the Mississippi may remain 
turbid for many years. 

2. Velocity and Transportation. The velocity of a current depends 
not only on the gradient, but also on the volume of water involved. 
Thus of two streams having similar gradients and form of channel, the 
one having the larger volume of water will have the swifter current. 
It is also well known that the swifter a current, the larger and heavier 
the masses it can transport, (it has been found that a current running 
a fifth of a mile in an hour will carry fine clay; one running half a mile 
in an hour will transport sand; one of a mile an hour will roll along me- 
dium-sized gravel, and one of 2 miles an hour will sweep along pebbles 
the size of an egg. (Reduced to mathematical form it may be stated 
that the maximum size of particles that a stream can move varies as the 
sixth power of the velocity?) If the velocity of a stream be doubled it can 
move particles sixty-four times as large as before. With low velocities 
of less than a mile an hour and with fine particles this increase does 
not seem very striking. Thus sand grains may be a hundred, or a 
thousand fold, as large as those of fine silts or muds and require a doubled 
or trebled velocity to move them. But with increasing speeds of miles 
per hour the effect becomes very marked. This explains why rapid 
streams of 5 miles per hour are able to move small boulders, and sudden 
floods in narrow valleys, caused by torrential downpours of rain or the 
bursting of dams, are able to carry with them huge masses of earth and 
rocks, sweep away bridges and other structures, and cause great damage 
(Fig. 24) . When the St. Francis dam near Los Angeles gave way in 1928 
and flooded the valley below, huge blocks of concrete, one of them weigh- 
ing 10,000 tons, were displaced by the escaping waters. 

3. Effect of Specific Gravity. The size of the particle that a stream 
of a given velocity is able to carry depends also on the specific gravity of 
the materials composing the particle. A familiar example is that a 
lead sinker is able to remain at rest on the bottom of a stream which 
carries away pebbles of an equal size. A practical application of im- 



portance is found in the fact that in placer gravels, from which gold is 
extracted, fine particles of the precious metal are mixed with vastly 
larger ones of sand and gravel; the water, on account of the high spe- 
cific gravity of the gold, being able to transport it only with difficulty. 
Hence if the gold grains are relatively angular, that 'is, unworn, it is 
inferred that they cannot have been transported far from the rocks 
which originally contained them and from which they were derived by 
erosion. When a prospector finds angular grains of gold, he searches 
the adjacent slopes for gold-bearing veins. The specific gravity of the 

Fig. 24. In times of flood a stream is able to carry masses and perform work vastly 
greater than under ordinary conditions, as here shown by the results of flooding. Manti 
Creek, Utah. (U. S. Geol. Surv.) 

great mass of material obtained by erosion and carried by rivers lies be- 
tween 2.5 and 3.0, that is, it is that much heavier than an equal volume 
of water. Also the principle that a body immersed in water loses weight 
equal to that of the water displaced, greatly aids the transporting 
power of the stream. 

(b) Transportation of Material in Solution. All river waters carry 
in solution salts of various kinds that have been leached from the 
rocks and soils of the country which they drain. Although in a measured 


volume of what we call fresh water the amount may seem relatively very 
small, in the aggregate the weight of material thus dissolved from the 
land and carried into the sea is enormous. It has been estimated that 
nearly 2,735,000,000 metric tons 1 of solid substances are thus annually 
transported into the oceans. The Mississippi carries about 136,000,000 
tons, the Connecticut, a small river, 1,000,000, the Danube over 
22,000,000, the Nile nearly 17,000,000. In the Mississippi the amount 
carried in solution is more than a third as large as that carried in me- 
chanical suspension, the quantities being 

340,500,000 tons in suspension; 
136,400,000 tons in solution. 

The most important of the substances thus dissolved and transported 
are calcium and magnesium carbonates (CaC0 3 and MgC0 3 ) ; calcium, 
sodium, and potassium sulphates (CaSCU, Na 2 S0 4 , and K 2 S0 4 ) ; sodium 
chloride (NaCl); and silica (Si0 2 ). In humid regions the abundant 
vegetation by its decay generates carbonic acid, and by the aid of this 
the percolating waters dissolve carbonates from the rocks. Hence in 
humid regions the waters of rivers like the Potomac and the Delaware, 
have chiefly carbonates in solution; in arid regions where vegetation is 
sparse or wanting, the waters contain mostly sulphates and chlorides, 
as in the Colorado and the Rio Grande. 

(c) Transportation on the River Bed; Traction. In addition to the 
material carried in suspension and solution a considerable part of the 
river's load is pushed or rolled along the bottom. What proportion of the 
whole this may be cannot be accurately determined; it has been thought 
that in the case of some rivers it is greater than the amount carried in 
suspension. From studies made on the Mississippi, it is roughly in- 
ferred that of the material which it carries into the Gulf of Mexico 
about 10 per cent or more consists of coarser debris moved along the 
bottom. It is obvious that, other things being equal, the steeper the 
gradient a river has the larger will be the amount of the material so 

By observation of experiments in troughs it has been found that 
the amount of material moved by actual sliding or rolling of the particles 
along the bottom of a stream is much smaller than the amount which 
progresses by a series of short leaps. Near the bottom of every stream 
there is a zone filled with grains moving in this fashion (saltation = 
jumping), and above this is the material in suspension. The particles 
urged forward by sliding, rolling, and saltation are said to be moved by 

1 Metric ton = 1000 kilograms = 2204 pounds. 


stream traction. The amount carried b}" traction, compared with 
suspension, varies with the swiftness of the current and size of the 

Manner of Transportation. In considering the manner in which 
material is carried one must recall that it is only in swift streams and the 
upper rapid tributaries of great rivers that boulders and coarse gravel 
are moved, especially in times of flood. As one goes down a great river 
the size of the material steadily grows less with diminishing gradient. 
This is seen not only in the matter in suspension, but on the bars and 
beaches where it is temporarily deposited. Finally, in those rivers which 
wander through wide plains before they reach the sea, only the finest 
sands, silts, and clays are discharged into the ocean, and no coarse ma- 
terial is seen, except chance pebbles and boulders that have been floated 
downstream in masses of river ice or among the roots of drifting trees. 

Nor is the journey a steady or uninterrupted one. The gradient 
changes from place to place and with it the velocity and transporting 
power. Material carried down one reach is deposited at the foot of it, 
while at the head of the next, rapid erosion is cutting the channel head- 
ward and material is thus again set in motion. Matter dropped during 
a season when the current is slack is seized and again hurried forward 
with the renewed strength that comes in times of flood. Thus, with 
many waits and pauses, and growing finer by attrition, the mass of ma- 
terial upon which the river works is urged ever forward and onward 
down stream. 


It has been estimated that the amount of material in suspension, in 
solution, and rolled on the bottom, discharged each year into the Gulf 
of Mexico, if gathered together would form a right-angled prism with a 
base 1 mile square and a height of 250 feet. If we reckon the whole 
basin of the Mississippi and its tributaries as covering 1,265,000 square 
miles, and consider only the material in suspension and solution, it can 
be calculated from the given data that the entire basin is being lowered 
at the average rate of 1 foot in 6000 years. An estimate for the whole 
United States, based on measurements made on its rivers, is about 1 
foot in 9000 years. The actual rate is probably greater, because the 
amount moved by traction is not included; and the two rivers, the 
Mississippi and the Colorado, which together transport about 80 per 
cent of the total material taken from the United States each year and 
delivered in suspension into the sea, are also those which must move 
the most by traction. Older estimates for the Mississippi basin have 
been as low as 1 foot in 4000 years, or even less. The rate for its basin 


must be faster than that of the United States as a whole, because certain 
large desert areas contribute very little to the annual run-off. 

The gradual lowering of a land surface through a long period of time is 
referred to as denudation. From the foregoing statements it is obvious 
that we cannot estimate the rate of denudation with any accuracy, but 
the results are of interest and importance because they indicate the 
order of magnitude of the figures concerned. We may say, with some 
confidence, that the area of the United States is being lowered at a rate 
of 1 foot in from 5000 to 10,000 years, and probably between 7000 and 
9000, and where rock many thousands of feet thick has been removed by 
denudation, we get some notion of the immensely long periods of time 
involved in the process. 

Other rivers, according to circumstances, have given different figures. 
Thus it has been calculated that the Ganges erodes its basin at the rate 
of 1 foot in about 1750 years. But its basin culminates against the 
loftiest mountains in the world and the river has a proportionately rapid 
descent and erosive power. The basin is also subject during part of the 
year to a very heavy rainfall and great floods. Consequently the rate 
is far greater than the average. On the other hand desert regions, like 
those in central Asia or the Sahara in Africa, with very little rainfall, 
are eroded with great slowness, the chief agent of transport being the 
wind. The average height of North America above the sea has been 
roughly estimated as 2000 feet; at the rate of 1 foot in 7500 years it 
would take 15,000,000 years to reduce it to sea level; but as erosive 
processes (excepting solution) go on more and more slowly as the slope 
is reduced, this time in reality would be enormously lengthened out. 

So far in the study of rivers we have considered the destructive, ero- 
sional work they perform work done chiefly in their upper reaches 
and seen in the valleys they excavate in the higher lands. Some rivers 
have a swift course through elevated tracts of country to the sea, their 
work is cut short when they enter it, and they deposit their load at once; 
but many, and especially the larger rivers of the world, descend into 
wide lowlands, through which with steady current they wind to their 
journey's end. In these lowlands, and at the rivers' mouths, the work 
done is different from that in the upper reaches; it is largely constructive, 
rather than destructive, and consists mainly in the deposition and slow 
shifting of the burden assumed through erosion in the higher part of the 
course. However, the work of cutting, especially against the valley 
sides, proceeds along with the work of deposition. It is not possible to 
separate the two processes in a comprehensive discussion. 



Meanders. As a stream reaches the middle and lower stretches of 
its valley its velocity constantly diminishes owing to decreasing gradient 
(Fig. 22). In consequence, it is easily turned by obstacles that it lacks 
the power to remove. Any deviation from a straight course throws the 

Fig. 25. Section of a stream channel 
from A to B shows it to have the profile 
seen above, the deepest part lying close in 
to the bank at A. See Fig. 26, A and B. 

Fig. 26. The formation of meanders 
and ox-bows. 

current against one of the stream's banks. Some erosion of the bank 
occurs, and the current is thrown against the opposite bank (as in A, 
Figs. 25 and 26) where again erosion takes place. Symmetrical curves 
are thus formed, and a continuation of the process increases them. 
Meanwhile the current is slackened at B (Fig. 26), deposition occurs 

Fig. 27. Meanders of a stream in nearly flat region. Trout Creek, Yellowstone Park. 

(U. S. Geol. Surv.) 

there, and the inside of the curve is built out as the outside is cut away. 
In this manner the arcs of the curves become more and more pronounced 
as CC and DD. Eventually a loop, as at E, is cut through, leaving an 
island in the stream, the main current takes the shortest route FF, and 
the entrances to the abandoned channel are quickly silted up, leaving a 


shallow crescentic lake. The symmetrical curves are called meanders 
(from the River Meander in Asia Minor) and the resulting lakes, 
oxbow lakes. Meanders are roughly proportional in size (o the size 
of the stream with which they are associated. The arcs in small streams 
commonly have circumferences of only a few score feet, but those of the 
Mississippi may be 20 miles or more in length. 

Lateral cutting is necessarily more effective against those portions 
of the meander curves which face upstream than against those which 
face downstream; hence each meander loop tends to shift slowly down- 
stream during its existence. Thus a constant succession of meanders 
moves almost imperceptibly mouthward past any given point in a stream 
valley. This steady shift downstream is called sweep. 

Alluvial Flats ; Lateral Planation. As a meandering stream wanders 
from side to side in its valley, it impinges against the valley bluffs from 

time to time (A, Fig. 28), cuts 
them down by undermining, and 
carries the material away. By 
continuation of this process the 
originally narrow valley is wid- 
ened. The work is known as 
lateral planation. Furthermore, 
by continuous deposition on the 

inside bank of ever ^ meander 

curve, the stream eventually cov- 
ers the entire floor of its valley with material dropped from its load. 
Whether the component material (river alluvium) is coarse gravel or fine 
silt, the resulting narrow plain of deposition is called an alluvial flat. 

Flood Plains ; Natural Levees. As the gradient of a stream steadily 
lessens, and as lateral planation increases, the sediment dropped is 
spread continuously along the valley bottom, building up great flats. 
In the spring, melting snows and heavy rains pour a vast volume of 
water into the tributaries of the stream. This volume, concentrated 
in the main stream, raises the water level until it overtops the channel 
and floods the bordering flats. The water which has been moving 
swiftly through a deep channel, with friction at a minimum, thus sud- 
denly forced out on to a shallow flat, is quickly and almost completely 
checked. Concentrated deposition therefore takes place along the 
immediate borders of the flooded channel, grading outward away from 
the stream into much thinner deposits. In consequence, low ridges 
are built up paralleling and confining the stream channel; and these 
remain after the flood has subsided and the stream has shrunk to its 
old dimensions. They are known as natural levees (Fig. 29). Beyond 



them the land is low and usually swampy. The whole broad flat, 
including both the natural levees and the flanking lower lands, is called 
the flood plain. The flood plain of the Mississippi covers an area of 
about 30,000 square miles, and a large portion of it consists of extensive 

With the subsidence of the flood, the decrease in volume of water 
flowing through the channel not only causes the stream to cease to 
corrade (i.e., scour out its channel); it may cause it actually to deposit. 
Thus during low water the stream silts up its channel with sediment, 
In this way the stream may gradually come to flow at a level higher than 
that of the flats beyond the levees, being restrained only by the levees 
from deserting its course and occupying the flats. This is the situation 
indicated in Fig. 29 and is moreover the condition in the lower Missis- 
sippi when in normal flood. This has become a serious matter from the 

Fig. 29. Flood plain with natural levees built by a large stream. 

point of view of floods, with the gradual settlement and agricultural 
development of the low, ill-drained areas back of the levees. 

Artificial Levees; Floods. Before the settlement of the lower 
Mississippi Valley, the river was in the habit of overflowing its banks 
periodically, the floodwater reaching the sea through the low flats 
beyond the natural levees. As early as 100 years ago attempts were 
made to confine the flooded river to its channel, in order to reclaim for 
agricultural purposes the adjacent rich and fertile alluvial lowlands. 
These attempts took the form of artificial levees which were built upon 
the natural ones, thus deepening and heightening the channel at the same 
time. If a given amount of water is confined to a narrowed course and 
thus prevented from normal spreading, its channel must necessarily be 
deepened. A part of this added depth is attained locally by increased 
erosion of the bottom, but much of it is attained by rise of the water 
surface. Hence whenever a new levee is put in, all those already in 
existence must be raised. The first levee, built at New Orleans, was 
4 feet high. Today the average height is more than 13 feet. In 1902 
there were 1300 miles of levees along the Mississippi; in 1927 there were 



2500. The river level is gradually rising higher above the adjacent 
basins, and when floods do occur, they are correspondingly more de- 
structive, as witness the disastrous flood of 1927. The river has been 
excluded from the flood plain, a part of its natural domain, and when it 
overflows its artificially restricted channel it is merely seeking its age- 
long rights. The building of levees is therefore not in itself an adequate 
method of preventing floods. The time is approaching when some ad- 
ditional means of combating floods will have to be adopted. 

The scheme of building levees has been followed in many parts of the 
world. In the flood plain of the Po .the river bed is above the house- 
tops. Breaks in the levees on the Hoangho at various times have re- 
sulted in enormous loss of life. In the flood of 1887 the Hoangho, 
called " China's Sorrow," drowned considerably more than a million 

Deltas. When the current at the mouth of a stream is checked by 
a body of standing water such as a lake or the ocean, its load is promptly 

Fig. 30. Ideal plan and section of a delta. Plan shows delta fingers and distribu- 
taries. Section shows thick, steeply dipping foreset beds overlain and underlain by 
thinner topset and bottomset beds, respectively. Relation of topset to foreset beds 
indicates sinking of the land as the delta grew. 

dropped on the bottom, building an embankment with a front which 
grows outward like a railroad or highway fill in process of construction 
across a valley. As the deposit is built up close to the water surface, the 
flood plain usually encroaches upon it from upstream and gradually 
covers it so that it is built above water in a crudely triangular shape with 
one apex pointing upstream. From this shape, resembling the Greek 
letter, the deposit derives its name of delta. This name is applied re- 
gardless of whether the embankment remains submerged or whether the 
encroachment of the flood plain has raised it above the water surface. 
The bulk of the delta-forming material is dropped on the frontal 
slope of the growing embankment, forming thick foreset beds (Fig. 30). 


Some of the finest sediment however remains longer in suspension and is 
carried farther out and dropped as fine botiomset beds which thin outward 
away from the delta. Along the top, where erosion and deposition 
alternate as the stream current changes seasonally, thin horizontal 
topset beds are laid down. The thickness of the whole mass is in some 
cases very great. Borings put clown in Venice, which is built on the 
delta of the Po ; reached a depth of more than 500 feet without attaining 
the bottom of the delta beds. 

Whenever the river in flood overtops the natural levees, it has oppor- 
tunities to break through at certain points and to flow seaward through 
some new channel, leaving a diminished volume of water to escape 
through the old channel. Since this happens so frequently as to be the 
rule in the case of large rivers left to themselves, new outlets are broken 
through soon after they have been formed, and a branching system of 
distributaries grows up, giving shape to the delta. Thus a long stream 
ending in a delta is like a rope frayed at both ends, the strands at the 
upper end being represented by the tributaries and the shorter ones at 
the lower end by the distributaries. The branching system extending 
seaward is the skeleton of the growing delta; between the long arms lie 
shallow basins which gradually fill with sediment during floods and thus 
become low land. 

The shifting of the main channel through the development of new 
distributaries is strikingly illustrated by the case of the Hoangho, which 
for approximately 700 years prior to 1852 had discharged eastward into 
the Yellow Sea. In 1852 it broke its banks at a point more than 300 
miles above its mouth, formed a new channel northeastward across the 
great alluvial flats of the province of Shantung, and finally emptied 
into the gulf of Chihli, almost 300 miles north of its old mouth. The 
Hoangho has occupied this new course with minor distributaries since 
1852, and the old channel has largely dried up. 

The Mississippi attempted a similar change in April, 1890, breaking 
its banks at the Nita Crevasse 1 between New Orleans and Baton Rouge, 
at a point well over a hundred miles from its mouth. From here it 
flowed eastward through Lake Maurepas, Lake Pontchartrain, and 
Lake Borgne into Mississippi Sound, inundating a wide area, causing 
great damage, and halting railroad traffic for two months. The river 
at length resumed its old course, after having taught local engineers 
that in the vicinity of the Nita Crevasse the Mississippi was 
normally about 21 feet above the level of the bordering swamps 
and flats, 

1 The term crevasse in the region of the lower Mississippi refers to a break in a 



IN 1852 t 


IN 19QS 

Pig. 31. Growth of the Mississippi delta during 50 years. (After G. R. Putnam.) 



Rate of Delta Building. In this way the land at the mouths of large 
rivers is constantly being added to at the expense of the sea. The rate 
at which this advance of the land takes place is variable, depending 
upon such factors as((l) depth of water offshore, (2) volume of stream- 
carried sediment, and (3) power of waves and shore currents to sweep 
away the newly deposited material.^ It is estimated that the delta of 
the Mississippi is pushing forward into the Gulf at the rate of more than 
250 feet each year (Fig. 31), and the Po pushes into the Adriatic at 
nearly as great a rate. Since 400 B.C., the Rhone has been encroaching 
on the Mediterranean at the rate of about 36 feet annually, whereas the 
deltas of the Danube and the Nile are each growing only about 13 feet 
yearly. The obstructions to navigation caused by deposition at the 
mouths of the Mississippi have been successfully removed by the build- 


Fig. 32. Delta of the Nile, showing its form and distributaries. 

ing of extensions of the natural levees out into shallow water. These 
extensions, called jetties, confine the current, so that with increased 
scour it deepens its channel and at the same time carries its load of 
sediment out into deep water before dropping it. 

Size and Form of Deltas. It follows from their rapid growth that the 
deltas of great rivers form large areas of land. The Nile delta (Fig. 32) 
is nearly 100 miles long and 200 miles broad on its seaward front. The 
combined delta formed by the Ganges and Brahmaputra is 200 miles 



long and has an area of possibly 40,000 square miles. The Mississippi 
delta is likewise 200 miles long, but is much narrower, having an area of 
not much more than 12,000 square miles. The Po delta, already large 
at the beginning of the Christian Era, has increased by nearly 100 square 
miles since that time. 

No two large deltas have the same appearance because of the capri- 
cious changes in the distributaries of the rivers and because waves and 

shore currents erode the deltas at varying 
rates. The Mississippi delta with its long 
projections (Fig. 31) built out successively 
by changing distributaries, shows that 
deposition by the stream is dominant over 
erosion by waves and shore currents in 
the Gulf of Mexico. It is a typical ex- 
ample of the " lobate " type. The Tiber 
(Fig. 33) has a " cuspate " delta in which 
shore erosion seems to have the upper 
hand over stream deposition, while the 
Nile delta (Fig. 32) is intermediate be- 
tween the other two and approximates an 
" arcuate " type (Johnson). 

Stream Terraces. Many stream val- 
leys, particularly in their middle reaches, 
are bordered by bench-like flats the tops 
of which are higher than even the flood 
stages of the present streams. In many 
cases there are several of them in series, 
rising away from the river like two long 
flights of steps facing each other. In some 
valleys they are continuous for long dis- 
tances. Such flats, high and dry, are 
called stream terraces. Upon examination 
of these forms it soon appears that some 

are made of rock (rock terraces) whereas others apparently contain no 
bedrock but consist entirely of unconsolidated sands, gravels, and clays 
deposited by the streams when they were flowing at higher levels (flood- 
plain terraces'). 

Rock Terraces. These are of two kinds. (1) Rock terraces found 
usually in dry regions where valleys have been cut down through 
alternating layers of weak and resistant rocks. The resistant layers, less 
easily eroded as the stream cuts down, are left standing out in relief, 
while the weak layers are etched back (Fig. 34). The terraces in the 


33. Cuspate delta of the 
Tiber. (Johnson.) 



Grand Canyon are an excellent example. (2) Rock terraces caused by 
successive uplifts of the land. These are mentioned here only for the 

Fig. 34. Rock terraces caused by differential erosion of a series of strata of unequal 


sake of completeness and are discussed in connection with changes of 
level (Chapter XVII). 

Flood-plain Terraces. Flood-plain terraces, the most common 
type, are formed by a stream which has begun to degrade following a 

Fig. 35. Terraces on the middle Fraser River, B. C. (F. F. Osbome.) 

long period of aggradation in which a flood plain was built up. It is 
obvious that any increase in the gradient or volume of an aggrading 
stream or decrease in its load would cause it to begin to degrade and hence 
to leave terraces (Figs. 35 and 36). A sudden upward warping of the 



land while streams are flowing over it, increasing their gradient, is 
regarded as a common cause. Another cause not rare in glaciated re- 
gions is the melting away of glaciers and the consequent great diminu- 
tion of the load of the streams that drained them. Many terraces also 

Fig. 36. Formation of flood-plain terraces. A A, section of river-cut valley; B, alluvial 
deposits of river; tt, former flood plain, now forming terraces; c, new flood plain. 

are rock-defended. They are developed in this way: A stream of low 
gradient which has been slowly excavating the deposits in the bottom 
of its valley, and is meandering from side to side across its valley floor, 
encounters at numerous points the bedrock of the valley wall. De- 
flected by this unexpected obstacle, it swings away and is thus prevented 

Fig. 37. Old flood plain "defended" against undercutting by a rock outcrop at X, 
thus gradually forming the terrace T. Base of block (vertical ruling) is solid rock. 

from undercutting the remnant of the old flood plain above the protective 
rock outcrop, which remains as a terrace, gradually increasing in height 
as the stream continues to cut downward (Fig. 37). 

Alluvial Cones and Fans. When a swift tributary stream enters 
a wide and nearly level valley, the abrupt change in its gradient may 



cause it to deposit the greater part of its load on the valley floor. 
In this way a low semi-conical elevation is formed, radiating out from 
the mouth of the tributary (Fig. 38). Such forms are generally known 
as alluvial fans, but if steep they are sometimes called alluvial cones. 
In one sense they may be regarded as deltas formed on land, but they 
differ from deltas in having a sloping rather than a flat top, since their 
upbuilding is not controlled by the surface of a body of standing water. 

Fig. 38. Alluvial cone, made by a tributary to a larger stream. Stoughton, "Wis. 

(U. S. Geol. Surv.) 

An alluvial fan acquires its shape because of uniform distribution of 
debris over its surface by -the parent stream. The stream repeatedly 
silts up its channel, overflows, and forms new distributaries. Thus 
when one part of the fan is built up ? the stream shifts to another course 
(which of necessity is temporarily lower) and builds that up. In this 
way the entire surface is covered by the stream. 

Alluvial fans are not common in humid regions. Where present at 
all they are small, and are usually developed in regions of soft material 
easy to erode, such for example as glacial deposits. In many dry regions, 
on the other hand, deposits of this type are characteristically large, and 
form conspicuous features of arid landscapes. 

Structure of River Deposits ; Stratification. All deposits by rivers 
k are so laid down through the sorting activity of water, that they consist 
v-df distinct layers or beds of varying thickness. Usually these beds are 


very regularly parallel for rather limited distances. Deposits which 
exhibit this laminated, banded, or bedded appearance are said to be 
stratified, and the arrangement is caUed stratification. In contrast to 
strata deposited on the sea floor, stream-laid beds have many local ir- 
regularities, and exhibit other peculiarities that reflect conditions in a 
stream valley. Many exposed sections of old rock strata show strati- 
fication characteristic of rivers, and we are thus enabled to reconstruct 
the courses of former streams so old that their deposits have had time 
to be buried and slowly converted into solid rock. 


Development of Valleys. Any land surface on which rain falls sheds 
the run-off down the lowest routes to the sea. The run-off loosens the 
surface soil along these low channels, carries away the d<bris, and thus 
excavates gullies. And stream valleys are merely gullies grown big. 
The evolution of gullies into valleys takes place in this way: 

Gullies form where the run-off is concentrated. Concentration of 
run-off enlarges a gully by erosion on its side slopes and at its head. 
The larger the gully becomes, the more thoroughly it concentrates the 

run-off, and an endless chain of cause and 
effect is thereby set up. Erosion at its head 
causes the gully to lengthen headward, and 
the slope wash down its sides widens it after 

= ^__ . each successive rain. At the same time the 

Fig. 39. Section of a river vai- even more concentrated run-off through the 
r^nTaX&r^?t^ e rafoved bottom deepens the gully and lengthens it 
by weathering and rain wash; m0 uthward as well. Thus constantly under- 

river r, trenching downward. ^^ Q^g^^ the gully becomes first a 

ravine and then a valley. The normal cross section of a valley which 
is undergoing rapid erosion is that of a V (Fig. 39), because the river, 
occupying a relatively small space, is cutting .downward, while at the 
same time rain wash and gullying tend to broaden the trench the river 
makes, by washing down the material composing the valley walls. As 
already shown, as fast as this debris reaches the river, it is seized and 
carried away. The cross section of a valley depends then on the relative 
balance between two agencies, downeutting by the river and broadening 
by weathering and rain wash. Thus in a region where the gradient is 
steep, downeutting by the river proceeds much more rapidly than 
weathering and rain wash, and the valley will be deeply incised and 
have profiles approaching ara (Fig. 39). As time goes on and the river 
gradient is lessened the cutting by the river becomes slower and slower; 



weathering then becomes relatively more and more pronounced and the 
valley widens out as shown in trb (Fig. 39). 
Intermittent Streams and Permanent Streams. While in the gully 

stage, the valley is likely to carry water only during and after rains. 

Fig, 40. A valley in a youthful stage of its history. 

Geol. Surv.) 

Yellowstone River. (U. S. 

Under these conditions the stream is said to be intermittent. As erosion 
deepens the gully, however, more and more of the rain water that enters 
the ground directly has opportunity to emerge again in the gully sides, 
and to contribute to the flow of water on the gully bottom long after the 
parent rains have ceased. Eventually the gully (perhaps a valley now) 
is excavated to and below the level at which all openings in the rock are 


permanently filled with ground water (Chapter VI). From this time on 
the water seeps steadily and uninterruptedly into the valley bottom, and 
the resulting stream is said to be permanent. 

Baselevel. When a stream reaches the sea its velocity is checked, 
its load is deposited, it can cut downward no further, and consequently 
it can erode no more. The sea level therefore is a level below which 
streams cannot cut. 1 The level of the sea, projected inland as an imagi- 
nary plane below the surface of the land, is called baselevel because it is 
the ultimate base and goal of denudation by streams. It follows that 
a stream hastens its own downfall every time it removes a cubic yard of 
material from any part of its bed, because thus by just so much it lowers 
the gradient on which its velocity depends. Since corrasion and trans- 
portation depend on velocity, which in turn depends largely on slope, it 
is evident that the gradient of a stream is steepest in the early stages of 
its history and that it progressively decreases as baselevel is approached. 
In Fig. 22 the base line represents the baselevel toward which the gradi- 
ent ab is steadily being lowered, but which it can never reach. 

Grade. Since in those places where the gradient is lessened a stream 
tends to deposit, while erosion again sets in when the gradient increases, 
it follows that, as time goes on, a river proceeds to fill up the hollows and 
to cut away the projections in its bed and thus to establish a definite 
gradient. The gradient which the river seeks to establish is one at 
which, in each part of its course, the velocity is sufficient for the volume 
of water there present to transport its burden without erosion or deposi- 
tion; the stream is then said to be at grade. This does not mean that 
the gradient is necessarily uniform from source to sea; it may be rela- 
tively much steeper in the upper course, where the load consists of coarse 
debris and the volume of water is small, than in the lower part where 
the slope is gentle but the volume of water is large and the load con- 
sists of fine sediment. A heavily loaded stream, like the Platte (Fig. 
41), may become graded on a relatively steep slope, as compared with 
one, not fully loaded, which on such a slope would be ungraded and still 
degrading. The lower parts of great rivers such as the Mississippi 
become graded while, in their headwaters, cutting and deepening by 
erosion are still actively going on. 

Thus in summary we may say that a stream is at grade when its 
transporting power and the load given it to carry are about equal. 
It is aggrading when the load it has to carry exceeds its ability to trans- 
port. It is degrading when its ability to do work is in excess of the ma- 

1 The fact that certain rivers can and do erode their lower channels well below 
sea level does not invalidate the application of the baselevel principle to broad areas. 



Fig. 41. A heavily loaded river. Note the wide bed with many shallow Interlacing 
channels and very broad valley. North Platte River above Gering, Neb. (U S Geol 

Fig. 42. General view of Niagara Falls. 



terial to be carried, and the excess of energy is employed in deepening 
its channel. When the stream has reached grade it is said to have 
reached its "profile of equilibrium " the profile which permits a 
transporting power closely adjusted to the amount of waste to be carried. 
Relation of Tributaries to Main Streams. Examination of drain- 
age systems shows that in a vast majority of cases the tributaries of a 

river enter it at grade, i.e., at the 
same elevation as the main stream. 
They are thus said to be accordant. 
The reason for this is that as the 
main stream cuts down its bed, the 
resulting increased gradient which is 
given the tributaries enables them 
to keep pace in downcutting in spite 
of the smaller volume of water they 
contain. But this may increase the 
ratio of the trenching of the lateral 
valleys over their widening to a 
greater degree than in the main trunk 
valley and hence they grow propor- 
tionately narrower and steeper. Ex- 
amples are afforded by some of the 
tributaries of Colorado River. In 
their effort to keep accordant rela- 
tions with the main stream they have 
cut narrow slot-like canyons. In 
some cases, however, in the younger 
stages of normal valleys, small trib- 
utary streams, unable to keep up 
with a rapidly downcutting river, 
are obliged to cascade down the main 
valley walls. 

Falls and Rapids, Falls and 
rapids are common in the valleys of 
swift streams. The majority of 
them are the result of the unequal 

Fig. 43. Map of Niagara River and erosion of rock masses Composed of 

Falls. (After G.K. Gilbert.) both hard and soft layers. This is 

magnificently illustrated in the great cataract at Niagara. Niagara 
River, which drains the four upper Great Lakes, in its course of 36 miles 
from Lake Erie flows over a plateau which terminates near Lake Ontario 
in an escarpment more than 300 feet high. The plateau is capped by a 



resistant layer of limestone under which are soft, easily eroded shales. 
Originally the falls was situated at Lewiston near the mouth of the river, 
and falling over the escarpment had its full height at this point. These 
relations are shown in Fig. 43. By the gradual disintegration and under- 
mining of the softer underlying shale the harder limestone on top is left 
projecting as a lip over which the water falls (Fig. 44). From time to 
time this projecting rock, left un- 
supported and penetrated by joint 
cracks, also falls and is carried 
away. By this means the falls 
maintains itself and at the same 
time steadily moves upstream, 
leaving a deep gorge behind it. It 
is now 7 miles above its original 

The recession of Niagara Falls, 

and the rate at which it takes Fig . 44. - Section showing rock layers and 
place, is a matter of interest and cause of falls at Niagara. (After G. K. 
, -, ,i i^' j. T~ i 7 Gilbert.) N. L.. Niagara limestone with soft 

has been the subject of much study sbale below . C ; L ._ ^ linton limestone ^ 

because it gives an idea Of the shales and sandstones below. 1 inch = 300 

length of time involved in geologic a w " L ' = water level of pooL 
processes. Successive surveys made throughout a period of 50 years 
have shown that the falls is retreating at a rate that averages 5 feet per 
year. If this rate was maintained from the time the falls first began to be 
cut, the length of time involved in the cutting of the gorge below the falls 
(7 miles) would be 7000 years. This is a minimum estimate, but the 
problem is not so simple as this, since many factors, involving various 
changes in the river and in the volume of its water during the past must 
be taken into account. Some estimates which have considered these, 
factors run as high as 35,000 years. Although we do not know the length 
of time with even an approach to accuracy these estimates are of value 
in that they show it is to be reckoned in tens of thousands of years, not 
in hundreds, nor in millions. 

Many other famous falls are due to an arrangement of rocks similar 
to that at Niagara, such as the Falls of St. Anthony on the Mississippi 
at Minneapolis, and its tributary streams, which fall into the gorge 
below; Shoshone Falls on the Snake River in Idaho; those on the tribu- 
taries of the Columbia River, and many others. 

Falls are caused in other ways as well; by glaciers, by the accidental 
damming of streams back of lava flows and landslides, and by uplift 
of the land relative to the sea. But whatever their cause, falls cannot 
indefinitely persist. Increased velocity at their crests results in in- 



creased erosion; thus the falls are worn down more rapidly than the 
reaches above and below them; they pass into rapids and disappear as 
the streams reduce their valleys to grade. 

Potholes. Circular excavations worn in bedrock by whirling eddies 
are common in the beds of streams below falls and rapids. They are 
called potholes (Fig. 45) . If the conformation of the stream bed is such 
that an eddy persists in one place, the water whirls sand and gravel with 

Fig. 45. Potholes in granite. Tuolumne River, Cal. (U. S. Geol. Surv.) 

it, and this bores downward; although the material wears out in grind- 
ing, it is continually replaced by fresh debris, and so the process con- 
tinues. Potholes have diameters ranging from a few inches up to 50 
feet ; their depth may be even greater. They are of interest in that they 
indicate clearly the action of whirling water and, occurring not un- 
commonly in rock now far from any stream, they prove that at one time 
this rock was the bed of a rapid current. 


If the foregoing conclusions on stream behavior are true; if streams 
steadily enlarge their valleys, and if tributaries develop and enlarge their 
valleys, and if cutting by the whole system is limited downward by a 
baselevel, what will be the final result of long-continued erosion by streams 
on a given land mass? Obviously there can be but one answer: the 


wasting away of the land to a low, gently sloping surface from which the 
water must drain sluggishly to the sea. This process, the complete 
denudation of a land mass, of course involves an enormous amount of 
time, but the final result is inevitable. The series of changes involved 
m the complete reduction of a region to baselevel constitutes a cycle 
of erosion. The time required necessarily varies with varying circum- 
stances such as initial elevation above the sea, resistance to erosion of 
the underlying rock, and amount of rainfall and run-off. 

The rainfall factor exercises the chief control over the process of de- 
nudation by streams, by controlling both the rate of erosion and the 
places where at a given time erosion and deposition occur Thus in 
regions where rainfall is very slight, the streams dry up before they reach 
the sea, and all their debris is deposited inland. The stages of the erosion 
cycle under various types of climate must next be considered. 

Although the process governed by the cycle is continuous, it is divided 
for convenience into stages, as follows: initial stage, youth, maturity, 
old age. It must be remembered that each of these stages grades into 
the next, and that all are parts of one unbroken chain of events. In 
connection with the following account of a typical case, reference should 
be constantly made to Figs. 46-51. 

Initial Stage; Consequent Streams. An upwarped area of sea 
bottom, bearing on its surface initial irregularities, appears above the 
sea. For the sake of simplicity let the material composing the mass be 
broadly homogeneous, with only local variations in its resistance to 
erosion. Gullies develop (Fig. 46) and grow mouthward and headward 
under the control of gravity and the initial slopes. Streams whose 
development is thus controlled by original surface irregularities are 
called consequent streams. Adjacent gullies grow into one another on 
favorable slopes, forming connected chains. They gradually become 
ravines and then valleys. In the early stages of the cycle, the gradient 
is so steep that downcutting of the valley bottoms is dominant over 
slope wash on the valley sides; hence the valleys and gullies are steep- 
sided and sharply V-shaped. Any irregularity of slope or material in 
the side of a gully is apt to concentrate run-off and thus to develop a 
tributary. Since no gully is uniform in these respects, tributaries 
rapidly develop, all lengthening themselves headward from the parent 
gully. When the initial gullies have developed initial tributaries, they 
break up the continuity of the initial slopes, and the land area passes 
from the initial stage into youth. 


Youthful Stage. The land area is now drained by an integrated 
drainage system consisting of main streams developed from the preexist- 
ing initial slopes. These slopes were so arranged that more concen- 

_ 47 

46 " ^ " " " " " " " 

Figs. 46-51. Ideal cycle of stream erosion under a humid climate and in homogeneous 

Pig. 46. Initial stage, showing gullies developing wherever the run-off is concentrated. 

Fig. 47. Early youth, showing integration of main drainage lines and growth of 
the stronger at the expense of the weaker streams. 

Fig. 48. Later youth, showing the reduction of the initial surface to irregular flat- 
topped ridges. 

Fig. 49. Early maturity, showing dissection of divides into flowing slopes and de- 
velopment of alluvial flats. 

Fig. 50. Later maturity, shoVing decrease in relief, lowering of slopes and widening 
of valleys. 

Fig. 51. Old age, showing development of peneplain with monadnocks. 

tration of drainage took place along the lines of the infant streams B 
and C (Fig. 47) than along streams A and D. In other words, the sum 
total of depressions in the areas of B and C made those areas lower than 
those of A and D. The chain of cause and effect was thus set up most 



rapidly and thoroughly along B and C and these streams sent out more 
tributaries and grew headward more swiftly than did A and D. Briefly 
stated, a struggle for existence takes place among adjacent streams in 
the competition to grow big and therefore to absorb larger drainage 
areas. In this case, the streams (B and C) favored by the initial slopes 
win out and maintain their lead over their less favored and therefore 
weaker neighbors (A and D). 

Divides. The area of higher land between two valleys is called a 
divide. The divide between two parallel gullies of unequal size may be 
completely destroyed by the growth of the larger gully (Fig. 52). The 

Fig. 52. Lateral seizure of a small gully by a larger one. 

divide between the heads of two streams flowing in opposite directions 
is easily shifted laterally. Such a situation is developing at x (Fig. 47). 
Two streams, c and d, tributary respectively to B and to C, have worked 
headward toward each other, thus narrowing the broad divide which 


Fig. 53. Evolution of a shifting divide into a permanent divide. Greater erosion 
by stream c than by stream d results in lateral shifting of the divide through the distance 
xx"- as well as in lowering it. At this stage, erosion has become equal on both sides of the 
divide so that further downcutting results merely in lowering the (now permanent) divide 
to &, x*, etc. 

formerly separated them. Stream c has a shorter journey to the sea 
than has stream d; hence the gradient of c is steeper; hence its power 
to corrade and to transport is greater; hence it eats both downward and 
headward more rapidly than does d. This inequality in rate of erosion 
results not only in lowering the divide x but in shifting it away from c 
toward d (Fig. 53). When erosion by c is equaled by erosion by d the 


divide ceases to move laterally and further erosion can only lower it. 
The divide is then said to be permanent. 

The streams in the youthful stage have gradients as steep as the height 
of the land above sea level permits; and since they are energetically 
cutting into their valley bottoms, the valleys are V-shaped in cross 
section, with steep sides whose slope depends on the resistance of the 
rock composing them (Fig. 48). The valley courses are crooked with 
irregular bends, all of them determined by the initial irregularities of the 
land surface. Tributaries develop rapidly, their valleys working head- 
ward from the main streams like branches growing from the trunks of 
trees. In fact, so closely does the pattern of a stream system under 
these conditions resemble a tree with its branches and twigs, that it is 
called a dendritic pattern. Each tributary enters its main valley at a 
level with the main stream. Downcutting by the tributaries keeps pace 
with downcutting by the main streams, and thus the whole system is 
delicately balanced and adjusted throughout its extent. The crooks 
and bends given each stream in its initial stage deflect the currents from 
side to side of their valleys, but downcutting is so rapid that no stream 
remains at one level long enough to allow appreciable widening of its 
valley by lateral cutting. As the countless tributaries continue to dis- 
sect the initial surface, the broad. initial divides contract into narrow 
and irregular ridges. The time of youth is the time of scenic grandeur 
in a landscape. Deep gash-like valleys (Fig. 40) and canyons with 
foaming rapids, precipitous cliffs, and high ridges are characteristic. 
But the very force that sculptured these forms will inevitably destroy 

It is during the period of youth that landslides are common in regions 
where steep slopes and suitable climatic conditions are found. The 
sliding of great masses of rock material from the steep sides of mountains 
and canyons hastens materially the destruction of the land. 

Mature Stage. When the tributaries have worked headward so 
far that the narrow ridge-like divides have been dissected into short 
hills and spurs, and when the valley sides and the tributary heads have 
destroyed all of the initial surface by converting it into slopes, the land- 
scape takes on a wholly new aspect and the region is said to be mature 
(Fig. 49). The intricate network of drainage is complete and all the 
inter-stream areas have been carved into slopes. The main streams 
have cut downward far enough to decrease their own gradients appreci- 
ably. As decreasing gradient progressively decreases each stream's 
downcutting power, and thus causes it to linger at each successive level, 
the force of the current deflected from side to side of its valley begins to 
cut effectively and each valley is thereby widened. In this way valley 



widening increases proportionately as valley deepening decreases, by 
the process known as lateral planation. Long stretches of the mam 
streams are at grade as they wander over their newly developed valley 
flats, while the upland (initial) surface, so prominent during youth, has 
melted into slopes. Since downcutting has been greatly checked, the 
mantle of loose material on the valley sides is not swept away as rapidly 
as it is formed by the agents of weathering. It therefore accumulates, 
and moves slowly down the slopes under the influence of gravity (locally 
aided by slope wash, ground water, and frost action). In this way hoi- 

Fig. 54. The smooth curving profiles of maturity, showing their relation to accumu- 
lated mantle rock which masks the horizontal strata. Compare the valley shown in 
Fig. 34, in which the horizontal strata appear unmasked because of continued rapid down- 
cutting by the stream. 

lows in the slopes are filled in, irregularities are smoothed out, and the 
profiles of the valley sides are converted into smooth flowing curves 
(Fig. 54). It is for this reason that the period of maturity is the time of 
restful beauty in a landscape. The rugged splendor of youth has been 
modeled into sweeping mellow curves. 

Old Age Stage. 'Because of the low and ever-decreasing gradients 
of the main streams, the heights are now wasted by the steeper tribu- 
taries much more rapidly than the larger valleys can be cut down. 
The result is a gradual decrease in relief as the divides are worn down to 
lower and lower levels (Fig. 50). Lateral planation by the main streams 
widens the valleys and thus helps to cut away the adjacent higher land. 
The streams are sluggish, meandering widely. Natural levees bordered 
by swamps are developed in their lower courses. As the bottoms of the 
main valleys slowly approach baselevel, corrasion by the main streams 
gradually ceases, and only the divides, where the gradients are still 
appreciable, are notably lowered. In this. way the hills of maturity 
melt down into low elevations in old age, shedding the run-off feebly in 
sluggish streams. Erosion takes place more and more slowly. The 
last few feet of vertical cutting might require a longer time than the 
entire amount of preceding excavation. The resulting surface of low 


relief, very gently undulatory, is called a peneplain (" almost a plain," 
Fig. 51). The highlands have been brought low; the rocks that compose 
them have been carried bit by bit to the sea and have been there deposi- 
ted in beds as sediment. Only a few residuals of the former high land 
remain. Here and there isolated hills (Fig. 51) rise above the general 
surface, like islands above a sea. They are composed either of very 
resistant rock or of masses so far from the main streams that they only 
stubbornly allow themselves to be graded down to the general level of 
their surroundings. Such island masses are called monadnocks after 
Mount Monadnock in New Hampshire which rises in this manner above 
the level of the surrounding country. 

The vast plain of central Russia has been cited as a good example of 
a modern peneplain; it has been slightly raised and the rivers have been 
set at work again eroding. Ancient peneplains, which have been uplifted 
and then carved by the streams into tracts of hilly country, have been 
recognized in many places, such as southern New England, Pennsyl- 
vania, central Missouri, and the south of England. 

It must be clearly understood that the topographic terms youth, 
maturity, and old age do not refer to periods of years, or to any absolute 
age. They denote merely stages, defined by the amount of work done 
in proportion to the total amount of work involved in the cycle. Thus 
a region of very soft rocks might reach old age while an area of resistant 
rocks was still in youth as far as the amount of erosion accomplished is 
concerned. It follows that an extensive valley system might exist in 
various topographic stages in different localities, depending on supply 
of water and the varying nature of the underlying rocks. As a matter of 
fact this is true of the valleys of most large rivers. 

Effect of Vegetation on Erosion. In humid regions the surface is 
commonly covered with an almost continuous blanket of sod, supple- 
mented locally by brush and forest. This mat of vegetation occupies 
uplands, slopes, and valley bottoms to the very edges of the streams. 
Through the action of frost and through repeated saturation with ground 
water, aided by gravity, the surface soil creeps down the slopes, carrying 
the vegetation with it. The movement is so slow and imperceptible 
that the mat of sod is rarely breached. Fresh gullying is hindered for 
several reasons: (1) Because the mass of roots distributed through the 
soil, together with the mat of organic matter on the surface, holds the 
soil firmly together and enables it to resist the pressure of the moving 
water, (2) Because the mat of vegetation acting like a sponge absorbs 
the water and permits it to drain off so slowly that the erosive effect 
of sudden rushes of water after storms is prevented. (3) Likewise in 
springtime the rapid melting of the snow is hindered by forest shade. 


Such effects are of course most noticeable on steep slopes, among hills 
or mountains. The profiles and contours resulting from this essentially 
unbroken protective covering, especially in the mature stage of the cycle, 
are smooth and flowing, a series of beautiful curves (Fig. 289). It is 
moreover a noticeable fact that in forest-covered countries the flow of 
the streams is less irregular than in non-forested regions, and the stream 
waters are relatively clear. 

If the forest cover of a country is removed, erosion proceeds rapidly 
(Fig. 55), and in a variety of ways great damage may be done. The 
regulative action of the forests on erosion and the flow of rivers is a 

Fig. 55. After the removal of the forest cover the soil has been carried away so rapidly 
that the remaining trees have their roots exposed by the lowering of the surface. Southern 
Appalachians. (U. S. Forest Service.) 

matter of great importance, not only from the geologic standpoint, 
but as vitally affecting civilization. In some countries, of which parts 
of northern China and Spain might be selected as examples, the im- 
provident removal of the entire forest cover has reduced large areas, 
through displacement and loss of arable soil by erosion, to sterile wastes, 
subjected alternately to hot and baking droughts and sudden disastrous 
floods. Destruction' of the forests by fire may have a similar effect. 
Once destroyed, and the soil washed out, they may be restored only 
with great difficulty after long periods of time. Considerable areas in 
the southern United States have been much impoverished in this way. 
In places where density of population places a premium on all arable 
land, terracing of hill slopes to prevent erosion is much resorted to. 
The yearly loss of valuable soil is one of the great wastes of modern 
civilization that should be checked as much as possible; forests should 
be cultivated on all eminences and places not adapted to agriculture and , 
their cutting carefully governed, not alone for the timber they furnish,* 
but to prevent erosion and regulate the flow of streams. 




Vegetation and Erosion in Semiarid Regions. The importance of 
the vegetation common to humid regions has been outlined in the pre- 
ceding paragraphs. Let us turn now to drier regions where vegetation 
is less abundant. In semiarid regions (regions where the annual rain- 
fall is roughly 10 to 20 inches), such as in much of the Great Plains 
region east of the Rocky Mountains, conditions somewhat resemble 
those in the deforested areas of the more humid country farther east. 
Trees are rare, the sod mat is present but not generally strong, and the 

Fig. 56. - 

- Steep-walled "wash," or stream channel. The bank is 30 feet high and has 
been cut since 1880. (Long-well.) 

soil is loose, dry, and porous. Rainfall, moreover, is likely to occur in 
sudden bursts. The result is rapid run-off and hence rapid erosion. 
Gullying occurs wherever the soil is laid bare, as on cattle trails and in 
wheel ruts. The streams are subject to sudden and heavy floods, their 
waters are very muddy, and in months of little rainfall they are com- 
monly low or even dry. Because of the low rate of weathering of their 
side walls, coupled with rapid downcutting by their streams, the valleys 
are steep-sided, and in addition they may be flat-bottomed because of 
excessive deposition (Fig. 56). 

Effect of Resistant Rocks; Canyons and Gorges. During the 
youthful stage of the cycle in humid regions, the streams cut down much 
faster than their valleys are widened by weathering and slope wash; 
and so if the initial elevation of the land is great, canyons and gorges (deep 
narrow valleys) may develop. But it is equally true that in maturity 
these must melt into wide, open valleys as downcutting decreases. 


In semiarid regions, however, the rate of valley widening by weather- 
ing is so slow that canyons and gorges, once they are developed, 
persist for a longer time. The grandest example is the Grand Canyon 
in Arizona, one of the most impressive wonders of the world. It is 
more than 200 miles long, 10 miles wide at the top, and from 3000 to 
6000 feet deep. In general its cross section shows a broader upper can- 

Fig. 57. Ideal section across the Grand Canyon. (After Dutton.) aa, outer canyon 
walls; bb, inner gorge; 1 and 3, hard resistant beds;" 2 and 4, soft beds. Vertical scale 

yon within which lies a deeper inner gorge (Fig. 58). It is cut in 
nearly horizontal beds of rock of varying hardness. These rest on crys- 
talline rocks of almost uniform resistance, which in one stretch have 
themselves been trenched to a depth of 2000 feet in the inner gorge. 
The more resistant rock layers form gaunt cliffs whose talus slopes 
partly cover the softer beds, and whose outcrops form broad terraces and 
platforms. These effects, and the irregular cutting, carving, and re- 
cessing of the canyon walls through ravines and side valleys, have given 
rise to enormous and striking architectural forms (Fig. 58). Some of the 
masses thus carved out are themselves large mountains. The river is a 
swift, turbulent stream, heavily laden with silt, from 200 to 300 feet 
wide and 2400 feet above sea level at the Bright Angel trail, the place 
in Arizona where the canyon is ordinarily seen. The Colorado must be 
considered as a young river in respect to the character of its valley and 
the magnitude of the erosive task that it has yet to accomplish. 

Effect of Weak Rocks ; Badlands. When weak materials such as 
clays and shales are laid open to rain wash, they are rapidly carved into 
gullies. Striking examples of such erosion are to be seen along the 
rivers that drain the Great Plains region. These rivers, such as the 
Missouri and its tributaries, the Cheyenne and the Platte, in places 
run in valleys sunk a considerable distance below the general level of 
the country. The rock that forms the sides of the valleys consists for 
the most part of very soft, barely consolidated clays and sands, easily 
cut by ram wash and gullying. The result is that on either side of the 
stream from the bottom-land by the river to the bench-land forming the 
plain, there lies a gradually rising belt of country dissected in the most 
intricate fashion by systems of gullies, gulches, and ravines, with spurs, 
knobs, and sharo ridges separating them. Such tracts of country are 



l Col rad - View is mostl ^ of ' the ^ gorge; the 

wail of the upper broader canyon is seen in the distance. (U. S. Geol. SUIT.) 



Fig. 59. Effect of rain wash in beds of clay. Detail. Sioux Co., Neb. (U. S. Geol. 


jijk % 50. A "hoodoo" in Monument Park, Colo. Hard masses of ironstone in beds 

of soft sandstone have shielded the rock below them from erosion and have thus produced 



known as badla?ids } because of the difficulty experienced in traversing 

Sculptured Forms: Buttes and Mesas; Koodoos. It has been 
explained that weathering of rocks is not everywhere uniform. All 
parts are not equally accessible through cracks and fissures by the agen- 
cies that produce decay, and some parts may be harder and more resist- 
ant than others. Because of this want of uniformity, remnants of the 
more resistant material are left as projecting masses. These masses 
protect the softer rock below, and pillars are thus formed (Fig. 60). 
Where these forms exist on a small scale they are often referred to as 
" hoodoos." Large isolated masses of soft rock capped and thus pro- 
tected by hard layers are known in the western United States as buttes 
(Fig. 61). Some are of mountainous size. Very broad flat-topped 

i * **.* * ; *., .!* 

Fig. 61. Red Butte, Bell Ranch, New Mex. (U. S. Geol. Stirv.) 

features, plateau-like in form, are termed mesas (Spanish " table ;; ). 
Mesas are capped by layers of hard ro'ck, in many cases lava, which have 
protected the softer layers beneath; These features testify to the great 
amount of material carried away by erosion from around them. Both 
buttes and mesas are present in humid regions but their appearance is 
not usually striking, because weathering and slope-wash conceal their 

Summary. The influences wrought by slight rainfall and consequent 
scanty vegetation impose themselves on the process of erosion through- 
out the semiarid cycle. The drainage pattern remains the same as in a 
humid region save that the tributaries are scanty instead of numerous. 



Alluvial fans play a much more prominent part in the semiarid than in 
the humid landscape, but following their extensive development in the 
youthful stage they are gradually cut away as the land is lowered, and 
the resulting surface is a peneplain. 


Some authors classify arid regions roughly as regions that receive 
less than 10 inches of rainfall annually; others classif} 1 - them as areas in 
which the drainage does not reach the sea. On the latter basis, one- 

Fig. 62. Outline map of a portion of southern Arizona showing both through-flowing 
and interior drainage. The large east-west stream is the Gila River. Many of its po- 
tential tributaries never reach it, and most of them (shown by dashed lines) are inter- 
mittent in their flow. Compare with the area of similar size shown in Fig. 297 

quarter of the land area of the globe is arid, if exception be allowed for 
through-flowing streams like the Nile and the Colorado. These streams 
maintain themselves through arid tracts in spite of great evaporation and 
lack of many tributaries because their headwaters in distant mountains 
give them a large and steady supply. 

Chief among these areas of interior drainage are the Sahara, the Libyan 
Desert, and the Kalahari in Africa, parts of the great Basin and Range 


region between the Wasatch Mountains and the Sierra Nevada in the 
western United States, the desert of western Australia; certain basins 
high in the Andes, and wide areas in central and western Asia. All are 
alike in that the streams which drain them lose themselves in the in- 
terior (Fig. 62). But they differ in many respects. In some, streams 
are active agents of erosion and deposition ; while in those that receive 
the least rainfall, the wind seems to be the chief dynamic agent. The 
latter have not been thoroughly studied. Because of this, and because 
no such deserts exist in North America, the discussion which follows is 
based upon the deserts of the United States. 

Controlling Principles of Denudation in Arid Regions 

Several dynamic processes are important in arid regions. The process 
of mechanical weathering is universal and is not further discussed. The 
other processes are as follows: 

(A) Stream Erosion and Deposition. Rainfall over the desert 
ranges sends clown torrents of water through the dry gullies and gulches 
and out on to the plains below. The excessive load of loose weathered 
material acquired in the steep gulches is rapidly deposited in broad 
alluvial fans as the stream velocities are abruptly checked and as the 
streams lose volume through evaporation and sinking. Fans at the 
mouths of adjacent valleys coalesce and in time build up broad apron- 
like piedmont plains. The alluvial fan, comparatively rare in humid 
regions and common in semiarid regions, is the universal unit of stream 
deposition in regions of aridity. 

(B) Sheetflood Erosion and Deposition. In certain districts where 
cloudbursts are rare but exceptionally violent, the flood water debouch- ' 
ing from adjacent mountain valleys spreads out and coalesces on the 
lower slopes. Moving with a continuous creeping or rolling motion 
the flood sweeps all before it, the water front, two or three feet high, 
" curling over and breaking in a belt of foam like the surf on a beach." 
The transportation and deposition of waste by these sheetfloods is 
an important process even though the floods are comparatively 

(C) Mudflows. Another normal though infrequently operative 
agent of gradation in arid regions is the mudflow. This is a process 
intermediate betweep. a sheetflood and a landslide. It occurs only 
where earthy material becomes watersoaked on steep slopes after heavy 
rains, and moves downward and outward as a slippery mass (Fig. 63). 
It advances in waves, stopping when it becomes too viscous to flow and 
damming the water behind it until it liquefies again and proceeds, like 



an advancing flow of lava. Mudflows can carry boulders many feet in 
diameter. Observers have seen these great rocks bobbing " like corks 
in a surf." In the course of time, successive mudflows play a large 
part in erosion and deposition. 

Fig. 63. Margin of a fresh thin mudflow. East side of the Still water Range, Nevada. 


(D) Landslides. Landslides are important in some places where 
. slopes are steep, where rocks are jointed so as to form great heavy blocks 

of talus, and especially where impervious shales make a slippery base 
for the mass of debris. The most favorable conditions for land- 
slides are not found in arid regions, but locally they play a part in 

(E) Deflation. Deflation is the picking up and exporting of fine 
material by the wind. The rapid mechanical weathering and scanty 
vegetation in arid regions are factors favorable to deflation; and in 
some deserts, as in the Sahara where sandstorms are frequent and violent, 
deflation plays an important role in degradation. The numerous sand 
dunes indicate the temporary resting place of the material in transit, 
and ships in the South Atlantic testify to the amount of fine material 
blown into the sea by the prevailing winds. Deflation is not, however, 
of great importance in the American deserts, its effects being largely 
masked by the agents cited above. 


Outline of the Cycle 

(1) Initial Stage. The following brief outline, following the studies 
of Davis, is organized with reference to Figs. 64-68. The conditions 
considered are to a certain extent special, but general principles are 
illustrated. Let a land surface in an arid climate be warped up into 
sharp folds. With the beginning of uplift, consequent streams develop, 
carrying drainage from the new highlands down into the adjacent 
troughs (Fig. 64). They flow only after violent storms and are repre- 
sented at other times only by dry valleys. Most of them evaporate or 
sink beneath the surface before they reach the bottom of the trough, 
depositing their loads in rows of alluvial fans which flank the highlands. 
Growth and coalescence of the fans result in the narrowing of the inter- 
mont basins. The floors of these basins are called play as (Fig. 114), 
and the waters that reach them are impounded as shallow playa 
lakes, only to evaporate soon afterward. The broader term bolson 
(Spanish " purse," i.e., a pocket or basin without outlet) is commonly 
used to describe the whole waste-filled valley (Fig. 69) between two 
desert ranges. Figure 64 shows the initial surface, initial consequent 
valleys, and two playas in the downwarped troughs. 

(2) Youthful Stage. Instead of being increased as in the normal 
cycle, the relief is slowly diminished by the removal of waste from the 
highlands and its deposition on the lower slopes and in the playas. 
In this way two bolsons are developed (Fig. 65), their centers occupied 
by playas and their sides flanked by ragged mountain escarpments 
strongly dissected by steep gulches and valleys. The baselevel, in- 
stead of being single and fixed as in a humid climate, is multiple, being 
formed by the various bolson surfaces; and is moreover gradually rising 
as the deposits rise at the expense of the wasting highlands. With 
decreasing gradients, the mountain streams deposit farther and farther 
up their valleys. Thus the heads of the fans migrate slowly backward 
toward the mountain crests. All the streams are intermittent, flowing 
only after rains. A certain amount of deflation takes place, especially 
of the fine material on the playa flat. 

(3) Mature Stage. As the mountain divide between two bolsons 
is cut down and the bolson floors are concomitantly built up, the higher 
basin in time comes to drain downward across the old divide into the 
lower one. When this occurs, the mature stage is said to be reached, 
and drainage may pass from the upper basin to the lower in times of 
rain (Fig. 66). The upper basin thereby begins to be dissected by a 
consequent system of gullies working headward from the new channel, 
and the debris from their excavation is deposited as a great fan in the 




Figs. 64-68. Ideal cycle of stream erosion under an arid climate. (Compare Fies 

Fig. 64. Initial stage, showing development of alluvial fans and playas. 

Fig. 65. Youthful stage, showing decrease of relief as the bolsons rise by filling. 

Fig. 66. Mature stage, showing capture of the higher bolson by the lower one. 

Fig. 67. Later mature stage, showing dissection of the higher bolson, transfer of the 
waste to the lower, and the exposure of pediments. 

Fig. 68. Old age stage, showing disintegration of the drainage, low relief, and climax 
of wind action. 



lower basin, hastening its filling. The surface of the lower basin has 
now become the master baselevel controlling both its own streams and 
those of its neighbor. 

Full maturity is reached when the upper basin is so thoroughly dis- 
sected (usually into badlands, since they are composed of loose uncon- 

Fig. 69. Bolson between Desert Range and Sheep Range, Nevada. Mature stage 
of the cycle. Compare Figs. 66 and 67. (Longwell.) 

solidated deposits) that every part of it drains down into the master 
basin (Fig. 67). The initial highland surface is gone, and the moun- 
tains are cloaked ever more completely with waste. 

(4) Old Age Stage. With the lowering of the mountains the rains, 
infrequent at the outset, become even more rare since condensation 
decreases with decreasing relief. The whole process outlined above 
is correspondingly retarded, but the wind, less hampered by stream 
action, tends to erode hollows in the fine loose material of the basins. 
This breaks up the drainage pattern by forming hollows in various parts 
of the master bolson. The higher basin ; stripped of most of its earlier 
mantle of waste, is largely floored with bare rock planed to a platform or 
pediment by the debouching mountain streams (Figs. 67 and 68). The 
wind becomes an increasingly important agent of erosion, disintegrating 
the drainage still further, and slowly lowering by deflation the whole 
surface, which now resembles a plain more nearly than a pair of basins. 
Only the rock masses that are most resistant to weathering remain 
as monadnocks. Since the wind can work at will over the entire area, 
the surface is slowly worn down to essentially the same level (Fig. 68). 
This final floor is thinly veneered with waste and dotted with monad- 


This last stage does not exist in any known desert at the present time 
probably because sufficient time has not elapsed since the recent Glacial 
Period (apparently a time of almost universal humidity) to allow it to 
develop. Penck, a German geologist, argues that such a surface could 
be worn down well below sea level, providing the sea were held out by 
surrounding highlands. The ultimate baselevel must be the ground 
water surface (Chapter VI) below which the mantle rock would be mois- 
tened and deflation thereby stopped. 


1. The Geographical Cycle; by W. M. Davis. In Geographical Essays, Ginn & 
Co., Boston, 1909, pp. 249-278. 

2. River Terraces in New England; by W. M. Davis. Ibid, pp. 514-586. 

3. The Geology of the Henry Mountains; by G. K. Gilbert. 160 pages. Wash- 
ington, 1877, pp. 99-150. A classic work. 

4. Rate of Recession of Niagara Falls; by G. K. Gilbert. 31 pages. U. S. 
Geol. Survey, Bull. 306, 1907. 

5. Exploration of the Colorado River of the West and Its Tributaries; by J. W. 
Powell. Washington, 1875, pp. 149-214. An early classic. 

6. Rivers of North America; by I. C. Russell. 327 pages. Putnam, New York, 
1898. A popular discussion. 


Glaciers carry off the accumulated snowfall from the lands, as streams 
remove the surface waters. They are rivers of ice. Wherever the 
winter snow does not melt in the summer it accumulates, and where such 
accumulation continues, glaciers originate. The study of existing gla- 
ciers discloses the pronounced changes they effect upon the land, and the 
observation of similar effects upon lands where today there are no glaciers 
leads inevitably to the conclusion that glaciers must once have existed 
there. The present is used as a key to the past. It will be our purpose, 
then, to study first the existing glaciers; next we shall examine the 
geologic work they perform, and then reconstruct the glaciers of the past. 


Perpetual Snow Fields. On all the continents, with the exception 
of Australia, there are places where some of the winter snow remains 
unmelted from year to year, giving rise to perpetual snow fields. Such 
places are more numerous in high latitudes than in low, and perpetual 
snow fields are a familiar sight on lofty mountains, even on those that 
lie beneath the equator. The level above which snow is perpetual, is 
known as the snow line (Fig. 70). At the equator it lies from 15,000 
to 18,000 feet above sea level, in Mexico 14,000 feet, in Yellowstone Park 
10,000 to 11,000 feet, in southern Canada about 9000 feet, in southern 
Alaska about 5000 feet, in Greenland about 2000 feet, and in arctic 
America a few hundred feet. The height of the snow line is determined 
not only by altitude and latitude (temperature), but also by the annual 
precipitation, humidity, and location on the sunny or shady side of a 
mountain. For example, in Bolivia the snow line is 18,500 feet in eleva- 
tion on the dry western side of the Andes, and 16,000 feet on the moister 
eastern side, and in parts of Alaska and Siberia where the ground re- 
mains frozen throughout the year the mean annual temperature is low 
enough for perpetual snow fields, if there were sufficient snowfall to 
exceed wastage by evaporation. 

Perpetual snow fields exist on most of the lofty mountain chains of 
North and South America. They are widespread in the higher moun- 
tains of Europe, Asia, and New Zealand; smaller ones occur even in 




tropical Africa. The greatest fields of snow and ice are in Antarctica 
and Greenland. 

In southern Alaska, where the winds from the warm Japan current, 
heavily charged with moisture, rise over the high southern mountains, 
there is an unusually heavy snowfall, which may amount to 50 or 60 

Fig. 70. Snow line on Mt. Fairweather (15,330 feet), Alaska; from the Pacific Ocean. 
(Alaska Glacier Studies, Nat. Geog. Soc.) 

feet in a year. This is the greatest region of mountain glaciers in the 
world, both as to total area and the number of individual glaciers. 

Neve ; Change into Ice. The greatest accumulations of snow occur 
in those parts of the perpetual snow fields whose angle of slope is less 
than that at which snow will slide. They are called gathering grounds, 
or, in the case of mountain snow B fields, catchment basins (Fig. 71). They 
receive not only that snow which falls directly on them hut also that 
which slides off the steeper slopes above. Tinder its own weight the 
snow becomes compacted and at the same time changes in character. 
The loose, feathery, newly fallen snow soon assumes a granular texture 
like coarse sand, and resembles the hailstone type of snow such as we 
see in the spring, in the remnants of winter snowdrifts. It is called 
n&)&. The change takes place largely as the result of alternate thaw- 
ing and freezing at the surface, during which, the larger grains of snow 
grow at the expense of the smaller ones. Whole snow fields so trans- 
formed become neve fields or slopes. 

Beneath the surface, if the thickness of snow is great, the neve be- 
comes compacted as the air is excluded, and passes into dense ice. 
This ice is more or less distinctly stratified, or banded, due to successive 



snowfalls of somewhat different consistencies, or to the presence of wind- 
blown dust. 

Movement ; the Glacier Formed. If snow and ice continued in- 
definitely to accumulate in the gathering grounds, there would, even in 
the brief space of historical records, be a thickness on the Alps that would 
reach to perhaps twice the elevation of Mont Blanc. But before any 

Fig. 71. The gathering grounds of the snows. Snow and neve fields on Mt. McKinley 
(20,300 feet). Note the transfer of snow by a snowslide from the upper steeper slopes to 
the catchment basin below, (La Voy.) 

considerable thickness can accumulate the masses of ice spread slowly 
outward and downward. Movement commences, and glaciers are 
formed. Each glacier, in mountainous regions, is a live river of ice by 
means of which the excess snow and ice are drained off from the higher 
places to the lower. It moves slowly clown the valley, profoundly 
changing it on the way, to a place where eventually its front melts. 

The exact starting point of a mountain glacier is somewhat vague. 
Its upper limit is usually considered to be the snow line, but there is 
movement in the neve fields far above this level. In most glaciers 
there is a zone of prominent cracks, called the bergschrund, between the 
snow slopes and the glacier. 

Not every snow field gives rise to a glacier; some are too small to 
form more than a tract of neve that passes into ice beneath. Such 
patches are common in all high mountains, as, for example, in Colorado 
where the general height of the mountains and the amount of precipita- 
tion are not adequate to cause real glaciers. 



^As an intermediate stage between a snow field and a glacier, some large 
neve fields form at their lower ends ice masses that give evidence of some 
movement but do not project as ice tongues for any appreciable distance 
below the snow line. Such masses are called glacierets, hanging glaciers, 
or cliff glaciers. Most of the so-called glaciers of the Rocky Mountains 
and the Sierra Nevada Mountains of the United States are of this class 
(Fig. 72). 

Kg. 72. A glacieret. Shepard Glacier, Glacier National Park. (Alden.) 

Lower Limit of Glaciers. Glaciers disappear either by melting or 
by emptying into the sea. The distance to -which a mountain glacier 
descends below the snow line before being halted by melting, depends 
upon the balance between the forward movement and melting. It 
might be likened to the distance a rod of ice could be thrust into a furnace 
before being melted; this depends upon the size of the rod, the rapidity 
with which it is pushed forward, and the heat of the furnace. Thus, in 
warm regions, glaciers in general project but a short distance below the 
snow line; if a glacier is large, or flows with relative rapidity, it will 
extend farther than under the reverse conditions. As we go to higher 
latitudes, the glaciers reach lower elevations; in sub-arctic regions they 
actually reach the sea and break off to form icebergs. The lower limit 
is also influenced by special climatic conditions, for in moist regions, 
where there is abundant snowfall, the glaciers are larger, flow more 
rapidly, and therefore descend farther, than in dry regions at the same 

In the Alps and in Norway, glaciers project as far as 5000 feet below 
the snow line. In southern Alaska they reach sea level at about 55 N. 


latitude, and in southern Greenland at about 60; whereas in Norway 
at 70 N, they melt before reaching the sea. In southern New Zealand 
at latitude 45 S. ? glaciers descend into tropical forests, and in Chile, 
glaciers from the Andes reach the sea at about 47 S. 


Glaciers are classified according to their shape, size, and location, as: 
valley or alpine glaciers; piedmont glaciers; and ice caps. Enormous 
ice caps, called continental glaciers, overran the northern countries in 
the recent geologic past, but are known only indirectly through their 
effects on the land surface. An understanding of them will be made 
clearer by considering first the separate types of existing glaciers. 

Valley Glaciers. It is valley glaciers that are usually thought of 
when glaciers are mentioned, for they are the common kind the world 
over. They are fed by mountain snow fields and flow down existing 

Fig. 73. Typical valley glacier with branches; the Siachen Glacier, Himalaya Moun- 
tains. Moraines of earth are seen on its surface as dark bands. 

valleys. They have been likened to rivers because, like rivers, they 
follow the valley windings and spread from side to side, and commonly 
they have tributaries (Fig. 73) that join the master ice stream and 
swell its mass. Furthermore a glacier may split about islands and 
come together again downstream as an unbroken mass of ice. But the 
resemblance to rivers is largely superficial; differences are more pro- 
nounced than are similarities. 

There are about 2000 valley glaciers in the Alps. It was here that 
they were first studied, and that is why they are often called Alpine 
glaciers. Most of them are less than 2 miles in length; a few are from 
3 to 5 miles long, and one, the Great Aletsch, is nearly 10 miles long. 
Similar valley glaciers may be seen in other parts of Europe, in 
Norway, the Pyrenees, and the Carpathians. Magnificent glaciers, up 
to 30 miles in length, are to be found in the Himalayas, and other high 
ranges of Asia, except the Altai Mountains, furnish fine examples. 
Even under the equatorial sun of Africa there is a small glacier on the 



slopes of Kilimanjaro (20,000 feet). In the United States hanging or 
cliff glaciers are numerous, but true glaciers lie only on some of the lofty 
volcanic peaks in the west, such as Mounts Rainier, Shasta (Fig. 74), 
Hood, and Baker. Those on Mount Rainier, in Washington, attain a 
length of 7 miles. The coastal mountains of British Columbia contain 
many fine glaciers. But the grandest glacial region of the world is in 
southern Alaska, where valley glaciers exist in unknown thousands, or 

Fig. 74. Mt. Shasta and Shastina from the west. Note that long glaciers descend 
on the north side, but none on the south side. (U. S. Army Air Corps.) 

perhaps tens of thousands, nourished by great unexplored snow fields 
that mantle the lofty mountains. Some attain 50 miles in length and 
5 to 6 miles in width, and descend from heights as great as 18,000 feet 
above sea level. Scores of them enter the sea and give rise to icebergs. 
Piedmont Glaciers. The name implies a glacier at the foot of a 
mountain. It designates those glaciers that have descended from the 
mountains and have spread bulb-like upon the gentler sloping plains 
beneath. The expanded foot of the Rhone Glacier (Fig. 81) may be 
considered as the beginning of 'a piedmont glacier. Usually several 
glaciers coalesce to form a piedmont glacier. If a valley glacier be 
likened to a river, then a piedmont glacier may be compared in size to a 



lake. Such glaciers are not common, and are confined to high lati- 
tudes; the Malaspina and Bering glaciers of Alaska are examples. 

The great Malaspina Glacier lies at the base of the range that supports 
Mt. St. Elias (18,000 feet) and Mt. Logan (19,540 feet), and spreads 
over an area of 1500 square miles to the edge of the sea (Fig. 75). It 
has a gently rolling surface, broken by innumerable fissures, irregular 
hummocks, and piles of debris. Its borders are so mantled by dirt that 

Fig. 75. Model of Malaspina Glacier, Alaska. (Martin, Univ. of Wisconsin.) 

they support a dense forest which rests on 1000 to 1500 feet of ice. 
The melting ice nourishes several short rivers of large volume. 

Ice Caps. These are vast sheets of ice which, to carry the analogy 
further, resemble seas in size. The picture they present is that of an 
almost endless monotony of desolate, wind-swept ice, gently rolling, and 
with occasional mountains of rock, known as nunatdks } projecting through 
them like islands. At their borders they taper down to elongate lobes, 
or to tongues of ice that discharge into the ocean, forming icebergs. 

Only two large ice caps exist today, in Greenland and Antarctica; 
smaller ones occur in Iceland. The Greenland ice cap is about 1300 
miles long and has an area of about 715,000 square miles. It has now 
been crossed several times notably by the explorer Koch. He states 


that there are two great flattish domes from which ice spreads out in 
all directions. The northern dome reaches an altitude of nearly 10,000 
feet and the southern about 8500 feet. The thickness of the ice probably 
ranges from 2000 to 7000 feet, The edge of the ice is definitely known 
to be in motion locally, since it discharges into the sea. In the northern 
part of Greenland, however, where the snowfall is light, the ice is thin 
and stagnant. 

The Antarctic ice cap is thought to have an area of 5,000,000 square 
miles. Its interior has been partly explored by Shackleton, Amundsen, 
Scott, Byrd, and Wilkins. Its thickness is unknown but its surface 
reaches an altitude of about 12,000 feet. According to Scott, it pushes 
off the land out over the sea to form vast stretches of floating fields of ice, 
known as the " Great Ice Barrier." The huge tabular icebergs of the 
Antarctic break off from the barrier ice. 

The great continental glaciers of the Ice Age, which have now entirely 
disappeared, must have resembled the ice caps of Greenland and Antarc- 


That glaciers actually move had long been suspected but was not 
generally known until 1827, when Hugi built a hut on the Aar Glacier, and 
its change of position was observed. Since then many accurate measure- 
ments have shown not only the rate but also the nature of the movement. 
The movement is always downstream, even though a glacier is said to 
be retreating; a retreat of a glacier is of course not a bodily movement 
of the ice back towards its source, but simply a retreat in the position 
of the ic front owing to excess of melting over forward motion. 

Rate of Movement. Hugi's hut moved down the Aar Glacier a dis- 
tance of 4650 feet in 15 years; 44 years later it was found 7900 feet down 
the valley, indicating a rate of movement of from 6 to 10 inches per day. 
A similar long-time measurement came to light for the Glacier des 
Bossons on Mont Blanc. In 1820 three guides were buried beneath 
an avalanche on the mountain. It was predicted by Dr. Forbes in 
1858 that their bodies would be given up by the glacier 35 to 40 years 
after their burial. Just 41 years later the heads of the three guides, 
with some hands and clothing, appeared at the foot of the glacier, so 
well preserved that they were recognized by friends. The average rate 
of movement was 8 inches a day. Other measurements have been made 
by placing markers on the ice and noting the time they have taken to 
travel a measured distance. Such measurements in the Alps have shown 
that the rate of movement there seldom exceeds 1 to 2 feet per day. 
In Alaska measurements made on the Kennecott Glacier (25 miles long; 


1 to 4 miles wide) showed an average rate of movement of the central 
portion, over a period of 1 year, of 2| inches per day; whereas the 
Childs Glacier in 1916 moved 4 feet per day, and in 1910, 8 to 40 feet 
per day. The large Muir Glacier was found to have a motion, at its 
center, of 7 feet per day. Rates as high as 60 to 75 feet per day have 
been recorded in some of the tongues of ice that move toward the sea 
through narrow mountain passes from the Greenland ice cap. 

The rate of movement is influenced by several factors: it is greatest 
in those glaciers that have a large supply of ice; it increases with a steep 
valley slope and a smooth bed; it increases with a steep upper slope of 
the ice; it is greater in summer, when there is more water in the ice, 
than in winter; it increases when the temperature of the ice approaches 
the freezing point. The gradient, the amount and thickness of ice, 
the steepness of the upper slope, and the temperature, are thus the 
chief factors that affect the rate of movement. 

Differential Movement. An important discovery in regard to 
glacier motion is that the rate of movement is not the same in all parts 
of its mass. A glacier does not move as a whole by sliding down its 
bed, like a cake of ice off the roof of a house. It has been found that 
stakes driven in a straight row across the top of a glacier after a time 
become curved downstream, proving that the middle portion moves 
faster than the "sides. Similarly, vertical lines of pegs on the side of a 
glacier show that the top moves faster than the bottom. Also, it has 
been observed that the line of swiftest motion does not always lie in the 
middle of the glacier, but is sinuous. The movement of the ice is 
therefore differential; that is, some parts of it move faster than other 
parts. This conclusion is important to a proper understanding of the 
cause of glacier motion. 

Nature of Glacier Movement. The observation that glacier move- 
ment is differential led early observers to assume that the flow of glacier 
ice is similar to that of a stiff, viscous fluid, such as pitch or asphalt. 
This, however, cannot be the case; the viscosity is only 'apparent, and 
not real. In a fluid, the molecules are free to move readily ^in any di- 
rection as they do in water or pitch, but in a crystalline solid the mol- 
ecules or atoms are fixed in definite positions from which they can be 
moved only by great force. Ice is a true solid; its component particles 
are definitely arranged with respect to each other in a geometrical pat- 
tern. Since it is crystalline, it cannot flow by free movement of the 
molecules and therefore cannot exhibit the property of a fluid. Fur- 
thermore, a moving glacier may contain many fissures and since flowing 
fluids do not crack, ice cannot move as a liquid. Also, a glacier will 
not ^turn and flow up an empty tributary valley, even though the top 


of the ice may tower high above the tributary floor; the Kennecott 
Glacier forms a dam several hundred feet high across Hidden Creek 
tributary, but the ice does not flow into Hidden Creek valley. Were 
the ice a true fluid, it would spread out into the empty tributary 
valley. Ice must flow as a solid; it does so because it becomes 
plastic under stress, as does a ductile metal, and so resembles a viscous 

Explanation of Glacier Movement The mechanism of glacier 
movement is complex and several ideas have been advanced to account 
for it; but to discuss them in full is beyond the scope of this work. 
Certain factors that have been considered important are: (1) Melting 
of ice under pressure and refreezing when the pressure is removed; 
(2) deformation of ice crystals along certain gliding planes without 
destroying its crystalline structure; (3) rotation of granules of ice; 
(4) interchange of individual molecules between ice granules; (5) shear- 
ing or sliding of one mass of ice over another. No one of these factors 
by itself is thought to be the sole cause of glacier motion, but rather 
that two or more of them operate together to produce movement. 
For convenience, these individual factors may be discussed separately. 

Melting and Refreezing. When the water from warm-weather melt- 
ing at the top of the glacier descends into the ice where the tempera- 
ture is lower, and there freezes, the expansion upon freezing causes a push 
in the glacier which will be in the direction of least resistance, or down 
the valley; also during its travel before refreezing, the water tends to 
move down the valley as well as down into the glacier, thereby causing 
a transfer of some of the bulk of the glacier to a lower place. 

Melting and refreezing have also been considered to aid movement 
by allowing individual granules of ice to rotate down grade; under 
pressure, the granules will liquefy where they bear on each other; the 
minute films of water so formed will move short distances to places of 
less pressure, where they will refreeze. This is called regelation. Thus 
the films of water move, and the granules from which they came may 
rotate slightly in response to gravity. Minute movements are set up, 
and if similar movements are taking place throughout all the glacier, 
there may be a slow motion of the glacier down the valley. 

Exchange of Molecules. Up in the neve fields, snow changes into 
granules of crystalline ice, and the larger of these grow at the expense 
of the smaller. There is thought to be an interchange of molecules from 
one granule to the other, in which the molecules move from a place of 
higher to one of lower pressure, and this would be down the valley. 
Such molecular interchanges taking place throughout the whole glacier 
may -give rise to a movement of the whole glacier. 


Shearing. It has been noticed, particularly in the lower ends of 
glaciers where the ice granules are large and interlocked, that there are 
planes along which one part of the ice mass has sheared over a lower 
part. The rate of slipping of upper masses over lower masses has 
actually been measured. It has been thought that this slipping or 
shearing accounts for part of the movement of glaciers, particularly in 
the lower ends. 

Crystal Gliding. In some crystalline solids of which ice is one, the 
component particles (molecules or atoms) are less tightly held together in 
certain directions than in others. Under stress, movement takes place 
along such planes of weakness between layers of molecules, without de- 
stroying the crystalline character. It has been demonstrated that such 
gliding, as it is called, takes place in ice along certain planes (for ex- 
ample, parallel to the base of the hexagonal prism), and ice, therefore, 
can adjust itself to pressure by movement as a solid. If such gliding 
should occur wherever ice granules happen to come into the proper 
position, a movement may be imparted to the whole ice mass. 

Probably several of the factors mentioned above play a part in bring- 
ing about motion. In view of the observation that movement is faster 
in spring than in fall and winter, and increases when the temperature is 
near the freezing point, it seems probable that liquefaction and regelation, 
aided by molecular interchange, play an important part in inducing 
movement. However, it is well known that many cold crystalline metals 
also flow, when stressed, under conditions of temperature where melting 
and refreezing are excluded. One need only deform a piece of lead or 
copper to be aware how small a force is required to make these weak 
metals flow. Ice is also a weak solid, and the mechanism of its flow may 
be similar to that of other weak crystalline solids. Recent investigations 
with the flowage of metals prove that, as in ice crystals, gliding also takes 
place in metal crystals, and that in addition granulation of the in- 
dividual grains enables movement to take place. 


The surface of a glacier must not be thought of as an expanse of smooth 
ice; typically it is highly irregular, so that travel over it may be la- 
borious and hazardous. It is traversed by wide and deep cracks called 
crevasseSj and is covered in places by accumulated heaps of earth and 
rock fragments called moraines. Unequal melting also gives rise to a 
variety of relief forms that hinder the traveler. 

Crevasses. Among the most striking features of a glacier are the 
great yawning cracks or crevasses in its upper surface. They may be 



20 feet or more in width and up to 300 feet In depth, and may occur 
anywhere between the neve fields and the lower end. They are formed 
by tension acting on the brittle ice, and are induced where the ice rides 
over an uneven bed, or where one part moves faster than another part. 
Their abundance varies with the amount and the rate of the straining. 

Fig. 76. Ice fall of tlje Franz Josef Glacier, New Zealand. Aimer Glacier entering on 
left. (New Zealand GeoL Surv.) 

They are most prominent where a glacier flows over a sharp incline in its 
bed, giving rise to an ice fall, as in Fig. 76. Transverse crevasses usu- 
ally bow downstream owing to the faster movement of the glacier in its 
central part. Even a change of slope in the bed of only two or three 
degrees produces crevasses, a striking proof that ice is not a true vis- 
cous substance. Where a glacier emerges from a constricted to a wider 
part of a valley, longitudinal crevasses occur, because the glacier tends 
to spread sideways and the ice is pulled apart. These are particularly 



characteristic of the terminal lobe of a glacier, as in Fig. 81. Marginal 
crevasses commonly occur along the sides of a glacier. These usually 
point inward and upstream at an angle of about 45 to the course of the 
glacier. They are formed as a result of tension set up by the central 
part moving faster than the marginal parts. 

Crevasses, once formed, do not remain open indefinitely. Like the 
particles of water of a river that break over a fall and join again below, 
so the ice breaks over an irregularity in its bed and heals again into a 
solid mass by refreezing; all evidence of crevasses may be obliterated. 
Not uncommonly, however, the upper parts of crevasses are widened by 
melting so that slices of ice between parallel cracks become sharpened to 
thin blades and needles, called seracs. These may persist long after the 
lower parts of the crevasses are closed, and a considerable part of the 
glacier may present a maze of sharp ridges such as one meets in trying 

to cross the Mer de Glace. On 

! ~" " ""* ~~1 large valley glaciers, such as those 

of Alaska, these ridges may take 
the form of steep hillocks 100 to 
200 feet high. 

Moraines. In a river valley, 
the debris that is worn from its 
sides forms local talus slopes or is 
washed down into the stream and 
borne away. But in a valley oc- 
cupied by a glacier, this material 
descends upon the ice to form a 
marginal fringe of debris (Fig. 77) 
that is carried along slowly by the 
moving ice . Thus there are formed 
continuous bands of debris along 
the margins of the glacier; these 
are called lateral moraines (Fig. 78). 
Where tributary glaciers enter, 
medial moraines are formed (Fig. 
78). Glaciers may be seen with 
several medial ribbons of debris, 
each separated from the others by 
" white ice/' and the number pres- 
ent usually indicates the number 
of tributaries that have joined the master glacier. As the glacier reaches 
lower levels and undergoes surface melting, the ice beneath the debris 
melts less rapidly, so that commonly the moraines may be prominent 

Fig. 77. Slide rock descending from 
bluffs upon a small glacier (white fore- 
ground) to form a marginal moraine. The 
mine buildings are situated upon the edge 
of the marginal moraine. Kennecott, 
Alaska. (Bateman.) 



ridges up to 100 feet in height. The debris may slide down the sides 
of these ridges and so widen the morainal bands. 

The debris is not only on top of the ice (super glacial) but is also within 
the body of the glacier (englacial). Much of the latter is carried along 
near the bottom of the glacier. The engla- 
cial material is derived partly from debris 
that fell on the surface as the glacier was 
being built up; partly from material that 
fell into the crevasses; and partly from ma- 
terial that was scraped off the sides and 
bottom of the valley. The englacial mate- 
rial may become superglacial where rapid 
melting occurs near the end of the glacier, 
and, in combination with spread-out lateral 
and medial moraines, may cover the entire 
" snout " of the glacier. The transported 
material that is eventually dumped at the 
end of the glacier, in a confused mass of 
earth and stones, forms the terminal moraine. 

Glaciers are thus great transporters of 
debris; fine dust, and boulders of thousands 
of tons weight are carried with equal facility. 

Differential Melting. The presence of 
debris on the surface of a glacier causes 
uneven melting by the sun, and numerous 
irregularities result. Thin layers of dark Fig. 78. Plan of a valley gla- 
debris or dust on the glacier will absorb cier, showing tributaries and ter- 

minal lobe; It, lateral moraines 

more heat from the sun than does the adja- running into #, the terminal mo- 
cent White ice, and will become Warmed rmne; mm, medial moraines. As 
,',,,,., ,, , . melting progresses more and 
through and melt the ice beneath to form more material appears; s, exit of 

dust Wells Or "bath tubs. 7 ' If the debris *f glacial ^earn ; rr, valley train 

. of water-laid debns. 

layers are thick, they act as heat insula- 
tors; the ice beneath is protected from melting and prominent ridges 
or dirt cones are formed. Some large slabs of rock eventually cap 
ice pedestals (Fig. 79) and are known as glacier tables. 

Drainage. If one walks over a glacier on a warm day, he hears the 
murmur of innumerable rivulets of water that are formed from the 
active melting that goes on. Many join together to make streams so 
wide that they are difficult to jump across, and these may be seen to 
plunge down deep crevasses, commonly into circular shafts called 
moulins or mills, of their own making. And deep beneath, one may hear 
the subdued roar of larger subglacial streams that flow in ice tunnels 



^f&^ : i^J^^^^ 

Fig. 79. Glacier table mounted on ice pedestal; caused by differential melting. 


Fig. 80. Subglacial stream and ice cave. Morainal material above falls as the ice 
melts and helps to build the terminal moraine. Transported blocks fill the bed of the 
turbulent stream which carries away the finer earth and ground-up rock. Chamonix, 



that they have made. All along the sides, also, numerous trickles of 
ice water descend and commonly form marginal streams. In a warm 
summer many feet may be melted from the top of a glacier. On one 
small glacier in Alaska, telephone poles sunk 6 feet in the ice had to be 
reset twice during the sum- 

The subglacial streams 
emerge from caves at the 
front of the glacier (Fig. 80), 
or less commonly they 'spurt 
forth from large vertical 
holes and form the starting 
points of good sized rivers. 
The water is striking in ap- 
pearance, for it has a pecul- 
iar opaque grayish-white 
color that has given rise to 
the name glacier milk. This 
is due to the presence of fine 
unweathered rock material 
that has been ground from 
the glacier bed and is carried 
in suspension. This milky 
water may be traced far 
down the valley or out to 
sea, and its presence makes 
the traveler suspect that he 
is approaching a region of 
melting glaciers. 

Advance and Recession 
of Living Glaciers. Glaci- 
ers appear to advance and 
retreat in response to vary- 
ing climatic cycles of greater 
or less precipitation, or of 
sunshine and cloudiness. 
At present there seems to be 
a general condition of reces- 

Kg. 81, Two views of the Rhone Glacier, Swit- 
zerland; the upper one taken in 1870, the lower in 
1905, illustrate the retreat of the ice in 35 years. 
The older one shows the terminal lobe and longitu- 
dinal crevasses. 

sion, although there are at 

the same time minor oscillations of retreat and advance. In Europe and 
North America glaciers have been retreating, in general, for several dec- 
ades. The Muir Glacier in Alaska has retreated more than 8 miles in the 


last 25 years, and since 1887 the Illecillewaet Glacier in British Columbia 
has retreated at least 500 feet. The amount of retreat of the Rhone 
Glacier is shown in Fig. 81. In the Chamonix Valley in the Alps, how- 
ever, the glacial retreat from 1812 to 1910 amounted at most to only 
| mile. 

On the other hand, some individual glaciers have shown prominent 
advances. For example, all the glaciers on the Savoy side of the Mont 
Blanc chain advanced prominently between 1910 and 1920. The 
Grindelwald Glacier, following a long retreat between 1875 and 1900, 
had by 1924 advanced a considerable distance. In Alaska, a lobe of the 
Childs Glacier advanced 2000 feet in 1910 to 1911, receded for the next 
few years, and advanced a short distance again in 1925. These ad- 
vances appear to be only temporary oscillations, although in the case 
of the Alaskan glaciers the length of time over which observations have 
extended is too short to state definitely whether there is a general retreat 
or advance. 


Studies of living glaciers teach us that they erode, transport, and 
deposit materials. And, with the work of existing glaciers in mind, we 
are led to infer, when similar effects are observed elsewhere, that glaciers 
were widespread in parts of the northern and southern hemispheres where 
today there are none. This was an ancient continental glaciation. 
Existing glaciers and their work are to be seen chiefly in mountainous 
regions, but the results of continental glaciation are nearly everywhere 
present in northern regions, even in flattish areas. The work of the 
vanished continental glaciers can be interpreted properly only by an 
understanding of what has been accomplished by living glaciers, or the 
present remnants of them, in the mountainous regions. Consequently, 
our attention will first be directed to glacial work in mountainous re- 
gions, which we can see actually taking place or which has taken place 
so recently that to fill in the gap involves but little speculation. 


Valley glaciers by their erosion not only give rise to distinctive floor 
features and characteristic valleys, but also profoundly modify those 
parts of mountains that project above the glaciers. How these features 
are produced will become clear by an understanding of the nature of 
glacial erosion. 

Nature of Glacial Erosion. Glaciers erode by plucking and by 
abrasion. The first process is most active in the upper reaches of a valley 



glacier, but operates throughout its course; abrasion is effective wherever 
the glacier moves in a well-defined channel. 

In plucking, the glacier actually quarries out masses of rock, incor- 
porates them into itself, and carries them along. This is accomplished 
by water that trickles down into crevices in the rock, freezes, and springs 
out blocks of rock, or that freezes around blocks bounded by joints or 
crevices so that they become a part of the body of the glacier, or by the 
ice of the glacier being pressed into the cracks in the rock by its" own 

Fig. 82. A group of glacial cirques (amphitheaters) ; also a small valley glacier, and 
a glacial stream and lake. Coast Range, British Columbia. (Bateman.) 

weight. Then, as the glacier moves forward, the blocks are plucked out 
and dragged along with it. It is a powerful quarrying operation. 
This process is most effective at the edges of the n4ve slopes, where the 
bergschrund facilitates ingress of water by day and freezing by night. 
Here there is a constant quarrying inward and downward, and in the 
course of time this gives rise to bowl-shaped basins called amphitheaters 
or cirques (Fig. 82). Commonly these are cut somewhat deeper at the 
center than at the place of discharge, and when the glacier disappears, 
these depressions become filled with water (Fig. 83) and form some of 
the beautiful mountain lakes or tarns, such as Lake Louise in the 
Canadian Rockies. Such cirques are common features of high moun- 



tains in middle and high latitudes, and are evidence of the former exist- 
ence of glaciers. 

The process of plucking operates also on the sides and bottoms of a 
valley glacier, particularly where projections of rock extend into the ice 
and where the bedrock is much jointed. 

Dirt and stones frozen fast in the bottom and sides of the ice form a 
huge slow-moving rasp that grinds away or abrades the bedrock of the 
bottom and sides of the valley. As the cutting tools wear out, new ones 
are continually supplied by plucking and abrasion. This rasping 
power is enormously augmented by the great weight of the overlying ice. 

Fig. 83. Lion Lake and small tarn in glacial cirque south of Triple Divide peak, 
Sierra Nevada, Cal. (Matthes.) 

The fact that a glacier erodes the sides of the valley it occupies, no less 
than its bottom, makes its erosional work entirely different from that of a 
river. The shape of a glacial valley is thus quite different from that 
fashioned by a river, and is distinctive. 

The rate of erosion is influenced by various factors. It is obvious 
that the larger the glacier, the greater will be its erosive power, because 
of its great weight; consequently master glaciers erode their valleys 
more than do smaller tributary glaciers. A thin ice sheet may override 
soil without even removing it. Also, clear ice produces no abrasion; 
the greater the amount of debris in the ice, up to certain limits, the 
more it can grind the bedrock. Erosion is favored by fast movement of 
the glacier and by weak bedrock. Another factor is the firmness with 
which the teeth of the rasp are gripped by the ice; where the ice is 
soft and melting, as at the lower ends, it may not even remove loose 
soil. Plucking plays a small part if the bedrock is smooth and un- 



jointed, but becomes highly effective if the rock surface Is rough and 

In contrast to a river, a glacier can cut far below grade locally and 
scoop deep hollows in bedrock along its course. As a result, glaciers 
erode valley bottoms unevenly; pronounced deepenings of a glaciated 
valley occur in constricted parts, where the ice has been forced to move 
faster, or in places underlain by exceptionally weak rock. 

Results of Erosion. Results of glacial erosion are evident in bed- 
rock features, in the character of the valleys, and in superglacial effects. 

The Valley Floor. If one walks up a valley recently vacated by a 
glacier, he will notice that the bottom and sides are smoothed, rounded, 

Fig. 84. Glacial striae; scratches and groovings made by moving iee on limestone 
bedrock. Near Rochester, N. Y. 

and scored by scratches and grooves, called glacial striae, that trend 
in the direction of glacial flow (Fig. 84). They were caused by the 
rock-shod bottom ice that dragged heavily over the floor; the pebbles 
and boulders scratched and grooved, and the dust polished. The rock 
flour thus formed was swept away by the subglacial drainage and gave 
rise to the " glacier milk " previously referred to. As one ascends 
farther, he may have difficulty in climbing steep little steps that face 
him abruptly but taper off smoothly upstream. These result from the 
plucking of blocks from the downstream sides of jointed hummocks of 



bedrock. The traveler may have to circle around many ponds or pools 
that occupy gouged-out rock basins in the valley floor. If he glances 
down the valley, he may note that the floor is made up of smoothed, 
rounded rock masses, elongated in the direction of flow, which, from their 
resemblance to the backs of a flock of sheep, are called roches moutonnees 
(Fig. 85). Some valleys also have a prominent stepped profile to which 
Russell has given the apt terra " a cyclopean stairway/ 5 The presence 
of all these features is a clear indication of the former existence of a 

Fig. 85. Roches moutonnees near Mono Pass, Sierra Nevada, Calif. (Matthes.) 

glacier. But they are not confined to mountain glaciers alone, for simi- 
lar features have also been produced by continental glaciers. 

The Glacial Valley. The shape of the valley is also a distinctive 
feature of erosion by mountain glaciers. It has a characteristic U- 
shape, flaring at the top, with steep sides, and a broad bottom. (Fig. 86.) 
Commonly it is several thousand feet deep. No river erosion could pro- 
duce such a valley. A comparison of Figs. 86 and 88 (a) shows the con- 
trast between a glacial valley and a normal youthful river valley. This 
unusual "shape results from the fact that a glacier erodes the sides as 
well as the bottom of its valley, and cuts away the ends of interlocking 
spurs developed between the tributaries of a stream-eroded valley. 

The tributaries also are peculiar. They do not join the master stream 
at grade as they normally do in river valleys, but commonly enter hun- 
dreds, or even a thousand or more feet, above the main valley floor. 
The streams that flow along such hanging valleys, as they are called, when 
they reach the abrupt drop, cascade down to the main valley floor and 


give rise to some of the most beautiful waterfalls (Fig. 87), such as 
those of the Yosemite Valley in California, the Lauterbrunnen in Switzer- 
land, or the Yoho Valley 
in British Columbia. The 
hanging valleys originate 
because the large master 
glaciers erode their valleys 
more rapidly than do the 
smaller tributary glaciers. 
Consequently, the master 
valley becomes overdeep- 
ened by bottom erosion 
with respect to the tribu- 
tary valley and the latter 
is left hanging, or the mas- 
ter valley becomes over- 
widened by lateral erosion, 
causing the mouth of the 
tributary to retreat up- 

stream and to be left hang- 

Fig. 86. U-shaped glacial valley. Hodnett 
Lakes Valley, Yukon Territory. (Geol. Surv. of 

ing above the bottom of the main 
valley. Both processes operate 
simultaneously, but in the case 
of high hanging valleys, vertical 
cutting has probably been much 
more effective than lateral cutting. 
Even where hanging valleys are not 
present there commonly exists an 
abrupt change in slope between 
the steep glacial valley wall and 
the unglaciated country above. In 
Switzerland these " shoulders," as 
they are called, are usually 1000 
feet or so above the valley bottoms 
and are a favorite place for villages. 
In Alaska they reach heights of 
3000 feet or more above the valley 

Other distinctive features of gla- 
cial valleys are truncated or faceted spurs, They are spurs of normal 
river valley erosion that have had their " noses " cut off by the erosion of 
the valley glaciers. Their characteristic triangular form may be seen 

Fig. 87. A hanging valley and falls. 
Yoho Valley, British Columbia. 


in Fig. 88, (6) and (c). Thus U-shaped sections, cirques, rock basins, 
hanging tributary valleys, shoulders, and facetted spurs are characteris- 
tic erosional features of glaciated valleys. The changes effected in a 
normal river valley by glaciation are seen in Fig. 88. 

Fiords. Where overdeepening in glacial valleys along a coastal 
region has extended beneath sea level, or where the coast has been de- 
pressed, the sea occupies the valleys after the ice has vanished, and 
fiords result. Most of the fiords of Alaska, British Columbia, Labrador, 
Chile, and Norway probably have been formed in this way. They are 
essentially no different from any glacial valley except that their bottoms 
are covered by the sea instead of being occupied by a river. Depths of 
water of more than 5000 feet have been measured in Alaskan fiords, 
and of 4000 feet in the Sogne Fiord in Norway. The depth of the 
Alaskan fiords is thought to be due to the depression of the coast more 
than to submarine excavation. 

Superglacid Erosion. The higher parts of mountains that rise 
above the glaciers and are covered by perpetual snow-fields are also 
greatly affected by ice erosion. This is shown graphically in Fig. 88, 
(a) to (c). Parallel tributary glaciers in the process of widening their 
valley sap the sides of intervening ridges and leave jagged comb ridges 
or aretes. The headward gnawing in cirque formation causes cirques 
to approach each other from opposite sides of a ridge until they inter- 
sect, leaving jagged sawtooth pinnacles between them (Fig. 82). Or, 
three or four cirques disposed about a single mountain and eating head- 
ward, as on the right hand mountains in Fig. 88 (6), will coalesce until 
an irregular or pyramidal horn is all that remains of a once massive moun- 
tain. Such horns are typical of glaciated mountains; the famous 
Matterhorn in the Alps and Mount Assiniboine in the Canadian Rockies 
are outstanding examples. The erosive work above the glaciers is thus 
chiefly sapping and undermining, and the results are ragged pinnacled 
slopes whose precipitous character usually defies the mountain climber. 


The vast amount of material transported by glaciers must be deposited 
when the glaciers melt. Consequently different kinds of deposits are 
formed, depending upon the character of the material deposited, the 
manner of the deposition, and the shapes and positions of the resulting 
depositional features. The materials transported and deposited by a 
glacier are called glacial deposits, or glacial drift. But the drift is 
of two distinct varieties: one, which is dumped in heterogeneous fash- 
ion as the ice melts, and is unstratified, is called till } or boulder clay; 



(a) A mountain 
mass, normally eroded 
by weathering and 
running water and un- 
affected by glacial ac- 
tion. The valleys and 
ravines are V-shaped in 

(&) The same mass, 
strongly affected by 
glaciers which occupy 
the valleys. Note the 
rugged topography 
above the ice, pro- 
duced by weathering 
and frost. 

(c) The same mass 
after the retreat and 
melting of the ice. 
Note the nature of 
the topography, the 
trough-like form of the 
glaciated valleys, the 
amphitheaters, some 
with lakes, the hanging 
valleys, and the facet- 
ed spurs. (W. M. 


the other is material that is sorted out and stratified by water from the 
melting ice, and laid down to form glacio-fluvial deposits. 

There is much similarity in the deposits of mountain and continental 
glaciers. Consequently the descriptions given for mountain glacier 
deposits will apply also in part to those formed from continental glaciers. 

Unstratified Deposits 

The outstanding characteristic of glacial till is that it is unstratified 
and in this respect it differs from the nicely assorted water- and wind- 
laid sediments. Coarse boulders and fine sediments occur jumbled 

together just as they were 
dropped by the melting 
ice. Some of the pebbles 
and boulders are as un- 
worn and angular as when 
they were removed from 
their original positions. 
Those that were dragged 
against the bedrock on the 
bottom or side of the val- 
ley, however, have charac- 

Fig. 89. Facetted and striated pebble. Sierra . . , . ,, , 

Nevada, Calif. (W. D. Johnson.) ten StlC Smooth flat SUT- 

faces ox facets (Fig. 89) and 

are polished, striated or scratched. No other pebbles or boulders are 
quite like them; they are diagnostic of glaciation. 

Moraines. When a glacier melts, the debris contained in the ice 
is dropped as a till sheet or ground moraine over the valley bottom. 
Commonly it is not very thick, and much of the finer material may be 
washed away, leaving the boulders as the most conspicuous part of the 

The lateral moraines of the glacier are left along the sides of the valley 
as long ridges of till, some of which are several hundred feet in height. 
The name lateral moraines, used for these features on the sides of live 
glaciers, is retained for the ridges after the glacier has disappeared. 
In places the ridges are sharp-topped, because the material fell to its 
angle of repose as the ice melted. The medial moraines of the glacier 
less commonly persist as distinct medial ridges on the ground; they are 
likely to be washed away, owing to their central position in the valley. 
In favored places, however, parts of old medial moraines are well pre- 

The terminal moraine consists of jumbled masses of till, usually of 



irregular shape, extending across the valley. Commonly it has a genera! 
crescentic form, concave toward the glacier. It is composed of pebbles 
and boulders, facetted and angular, heterogeneously admixed with earthy 
materials. The piles range in height from a few tens to a few hundreds of 
feet. The higher ones usually indicate that the ice front was relatively 
stationary for a long time 
so that much debris was . 

deposited at one place. A 
glacier that retreats com- 
paratively fast leaves only 
a low, scattered terminal 
moraine. If the ice front 
in its retreat makes ex- 
tended halts at intervals, a 

. , .' 

series of recessional moraines - ** H 



90. Erratic block, Alaska boundary. 
(Coleman, after Melson.) 

be left. The terminal 
moraine is usually breached 
by streams and may be en- 
tirely removed by them. 
The surface of a typical 
terminal moraine is charac- 
terized by numerous round- 
ed mounds and small un- 
drained basins, with a hap- 
hazard distribution. This 
peculiar kind of surface, 

known as knob-and-basin topography, is caused by highly irregular de- 
position of the terminal moraine. This irregularity results from fre- 
quent minor fluctuations of the ice front and from variation in amount 
of debris delivered by the ice to any one point. Some of the basins in 
an abandoned terminal moraine form small ponds. 

Glacial Boulders, or Erratics. Another characteristic feature of a 
glaciated valley is the presence of scattered boulders, or erratics, of all 
shapes and sizes, many of which are foreign to the underlying rock; 
boulders of granite derived from the head of the valley may rest on 
limestone in the lower part of the valley. Commonly they are perched 
in insecure positions (perched boulders), or they may be so nicely poised 
that they can be rocked by the hand (rocking stones). They may be 
distinguished from somewhat similar boulders of decomposition by their 
dissimilarity in composition from the underlying rock. These erratics 
will be referred to again under continental glaciation. One may be 
seen in Fig. 90. 



Glado-fluvial Deposits or Stratified Drift 

The streams discharged by melting glaciers carry out boulders, 
pebbles, sand, and silt. The boulders, however, are quickly dropped; 
the remaining materials build up an outwash plain, or a valley train 
which may extend far down the valley. Nearest the glacier are the 
coarser pebbles, and farthest away are the sands or coarser silt; the 
finest rock flour of the milky water may be carried far beyond the valley 
train. The deposits are thus sorted put by the water and are stratified 

Fig. 91. Pitted outwash plain, derived from Hidden Glacier, Alaska. Part of ice 
near front of glacier shows in foreground. Note the braided streams, which are building 
up the plain. (U. S. Geol. Surv.) 

into beds of sands and pebbles. The latter are somewhat rounded by 
the friction they experience as they are rolled along. Across these 
gravelly plains, where the gradient is low, the streams pursue a braided 
course and are continually shifting their channels. They thus work 
over the material and search out the finer sands to carry them farther 
downstream, leaving behind extensive beds of well-sorted, clean pebbles, 
whose thickness may be scores of feet. As the outwash material ac- 
cumulates it may lap up over the end of a stagnant glacier, or bury large 
isolated blocks of ice. Later, when the ice melts, pits or kettle holes 
result, and the stratified deposits form a pitted outwash plain. (Fig. 



Continental glaciation on a far grander scale than that of mountain 
glaciation has taken place over the greater pail of high-latitude lands. 
The glaciers themselves have largely vanished; their former presence is 
inferred from the results they achieved, and their character is arrived 
at by inference and by analogy with existing ice caps. The nature of 
the evidence, however, in the light of our present knowledge is con- 
clusive. They overran regions where mountains are few and small, 
as well as some mountainous districts; and they are supposed to have 
been great flat domes of ice, so thick that they covered most of the lands. 
Their action, therefore, was different from thac of valley glaciers; they 
did not flow as tongues of concentrated erosive power in valleys, and 
therefore glacial valleys, hanging valleys, and cirques are not character- 
istic of their work; they flowed from centers that were hundreds of 
miles from their terminals, and hence carried some materials great dis- 
tances; they carried little or no debris on their surfaces, and as the} r had 
no side boundaries they had no lateral moraines. Their morainal ma- 
terial came largely from the underlying bedrock, and therefore much of 
it shows signs of wear; their melting took place over a wide front, with 
the result that their terminal moraines and outwash deposits are of 
vast extent; they overrode hill and vale, and hence much of their 
deposit is independent of topography. We shall first examine the signs 
they have left and then deduce their character. 


The Floor. The continental glaciers were powerful engines of erosion ; 
over wide areas they removed all soil and weathered rock, and plucked 
and abraded the floor, giving rise to vast stretches of gently rolling 
surfaces that display glacial striae, groovings, and roches moutonnees 
on fresh unweathered rock. These roches moutonnees have a smooth, 
gently sloping up-glacier side and a more abrupt down-glacier slope. 
The glaciers overrode hills and sculptured them into shapes similar 
to the smaller roches moutonnees. But toward the southern edge of the 
ice advance, pronounced erosional features are less conspicuous, al- 
though striae and grooves are abundant ; in places the weathered surface 
rock has not been entirely removed. Erosional features in the regions 
of continental glaeiation are, on the whole, less evident than depositional 

Glacial Lakes. Throughout the glaciated region of eastern North 
America are hundreds of thousands of lakes, large and small, that have 


resulted from glaciation. ' In the northern areas they have been formed 
chiefly by the flooding of rock basins scooped out by the ice. In the 
southern regions, a more important cause is the formation of depressions 
by irregularly deposited debris, or the deposition of moraines athwart 
valleys, forming dams that pond the drainage. So numerous and 
closely spaced are these glacial lakes that what would otherwise be a 
little known wilderness in northeastern North America is made accessible 
to travel by canoe. In Minnesota alone there are 10,000 lakes. Of the 
many beautiful lakes that lend charm to the scenery of the rolling coun- 
try of New England, the Adirondacks, Minnesota, Canada, or Norway, 
by far the greatest number have resulted from the work of glaciers. 
In the unglaciated country of the southern and central United States, 
lakes are rare or wanting. 


The continental glaciers carried great quantities of rock debris 
which later was dropped on the land or spread afar by the action of water. 
Because of their great size and distribution, they transported vastly 
greater amounts of debris than valley glaciers. Most of the material 
of the drift has not been moved more than 50 miles, although some of 
it has travelled hundreds of miles. This has been determined by the 
finding of materials far distant from their place of origin. For example, 
slabs of native copper that came from the great t copper lodes of the 
Keweenawan Peninsula of Michigan have been found as far south as 
Missouri; likewise, boulders of an unusual conglomerate containing 
reddish pebbles of jasper found throughout Ohio have been traced to a 
bed on the north shore of Georgian Bay, Canada. In northern Sweden, 
glacial boulders of copper ore that nad been transported by the ice were 
traced back in the direction of ice flow and resulted in the discovery, 
by electrical surveys in 1925, of Sweden's most important copper mine. 
Of unusual interest in this connection was the finding of several isolated 
diamonds of good quality in glacial deposits in Wisconsin, Michigan, 
Ohio, and Indiana. Their position in the deposits indicated clearly that 
they had been transported and deposited by the ice, but their source is 
as yet unknown. It is presumed to be somewhere to the northward in 


The disappearance of the continental glaciers left the vast regions 
overridden by them mantled with glacial debris of various lands and 
arrangements. These deposits have exerted a profound effect on the 


social and economic development of the countries with a glacial past, 
for the fertility or sterility of the soil, the drainage, and in part the 
topography,, have been conditioned by such deposits. 

As in the case of the existing glaciers previously discussed., deposits 
were formed of unsorted material dropped from the melting ice, and also 
by means of running water. Therefore both glacial till and glaeio- 
fluvial deposits are widespread in the glaciated countries. 

The Till or Unstratified Deposits 

The Till Sheet. When the ice melted, its load of debris was dropped 
as a sheet of till that overspread the land to a variable thickness ranging 
up to several hundred feet. It consists of old soil, ground-up rock, small 
fragments, pebbles, and boulders in heterogeneous disorder, and com- 
monly the adjacent boulders have no similarity in composition. (Fig. 
92.) Many of the pebbles and boulders are polished, scratched, and 


A "^^^^^^J-, 

Fig. 92. Drumlin near Newark, N. Y. (U. S. GeoL Surv.) 

facetted (Fig. 89). Glacial boulders are so common that one may pick 
them up at random in the glaciated regions; the many stone fences oi 
New England attest their abundance. In places they are so numerous 
as to render the soil untillable. It is this bouldery character of the til 1 
that has given rise to the term boulder day. The till sheet was depositec 
somewhat evenly over the bedrock, but variations in thickness imparl 
to the surface a gently rolling character, with low swells and shallow 
swales or depressions. In some localities there are smooth, oval-shapec 
hills of till, elongated in the direction of ice movement. These arc 



drumlins; it is not known definitely how they assumed their form and 
position (Fig. 92). They are especially common in New York, eastern 
Massachusetts, and eastern Wisconsin. In general, they average about 
a half-mile in length and 100 to 150 feet in height; and commonly they 
occur in clusters. 

Moraines. Near the margins of continental glaciation one may see 
belts a few hundred feet to several miles broad that look as though an 
army of excavators had dug a maze of depressions and stacked the debris 
in nearby dump piles. There is bewildering confusion in the haphazard 

Fig. 93. Glacial till, Bangor, Pa., consisting of unassorted boulders and fine materials. 

(Pa. Geol. Survey.) 

arrangement of pits and mounds. They are spaced without order. 
Examination of the material itself reveals glacial pebbles and the lack 
of sorting characteristic of till (Fig. 93). This is the work of the glacier, 
and these features are typical of its terminal moraine. In its extensive- 
ness, it bears little resemblance to the small terminal moraines of valley 
glaciers, for these large ones may be followed almost without break 
across states and countries. However, the peculiar knob-and-basm 
topography in the great moraines has the same general origin as in the 
analogous valley moraines; it is due to the irregular deposition of 
debris liberated from a fluctuating ice front. If the ice, for a period of 
time, melts about as rapidly as it advances, much material will be 
dropped in one place. As the ice front oscillates, and projecting lobes 
tod embayments change their position and shape, the frontal depo- 


sition is necessarily hummocky. The knobs may rise to 100 or 200 
feet above the basins, or there may be only a few feet difference in 
elevation. The depressions are commonly occupied by marshes or 
ponds that are numbered in the thousands. In places the terminal 
moraine consists only of a prominent riclge or a series of separate or over- 
lapping ridges. 

Every considerable halt of the ice front in its retreat gave rise to other 
moraines, called recessional moraines, and the unsteady retreat of the 
vanishing glaciers can be traced as surely as though they had been under 
observation. Long lobes of ice must have lagged behind the retreating 
main sheet, for in many places the terminal and recessional moraines 
extend parallel to the direction of movement, indicating deposition along 
the sides of such tongues. Kettle holes formed from the melting of 
residual blocks of ice are numerous in the vicinity of the terminal 

Erratics. Unlike the glacial boulders or erratics of mountain glaciers 
that are found only in valleys, those dropped by continental glaciers 

Fig. 94. Glacial erratic, a transported boulder of trap resting on sandstone; weight 
about 500 tons. New Haven, Conn. 

occur far and wide over hill and vale. It is not uncommon to see them 
on the very tops of high hills, in positions where they could not have 
fallen from higher ground or have been carried by streams. And when, 
in addition, they rest on till and differ in composition from the underly- 
ing rock, it is certain they were left by glaciers (Fig. 94). 

Glade-Fluvial Deposits 

The disgorging of torrents laden with sediments, as described for the 
small living glaciers, must also havje taken place on a vastly greater scale 
from the hundreds of miles of front of a continental glacier. This is 
not mere supposition, for the extensive glacio-fiuvial deposits that lie 



within and around the regions of continental glaciation abundantly 
prove it. However, a greater variety of water-laid deposits emanated 
from continental glaciers than from valley glaciers, and there is evi- 
dence that some were formed under the ice, and others near its edge, 
in addition to those that were laid down beyond the ice front. 

Outwash Plains. In front of the continental glaciers, valley trains 
such as those described in connection with existing valley glaciers formed 

Fig. 95. Stratified drift. Excavation for Yale Library, New Haven, 
Conn. (Longwell.) 

only where well-defined valleys were present. But valleys were infre- 
quent in the open country fronting the continental ice caps, and most of 
the water's burden was deposited to form great coalescing alluvial sheets, 
known as outwash plains or frontal aprons. Like the smaller valley 
trains, they are composed or water-sorted gravels and sands, built up 
to considerable depths by streams that rapidly became graded and 
braided. Some of the finer rock flour supplied the material to make 
the loess deposits of the Mississippi Valley and in parts of Europe. 
Southern Long Island was built up as a sandy outwash plain, as was also 
a large part of southern Ohio. With the retreat of the ice, outwash 
plains came to overlie earlier till deposits (Fig. 96). 
Some of the outwash plains are prominently pitted by large and small 


kettle holes, where the out-wash materials have slumped after the melting 
of stagnant ice masses that were partly or entirely burled by the accumu- 
lating sediments. Many of these kettles, some as deep as 100 feet, 
occur in the sandy outwash plain at New Haven, Connecticut. 

The sands and gravels of outwash plains are widely used as materials 
for road and building construction. 

Fig. 96. Water-laid glacial drift on unassorted till. Columbus, Ohio. 

Kames and Eskers. These are peculiar forms of deposits made by 
the sediment-laden streams from the melting ice. Kames are hummocky 
hills and ridges composed of stratified drift, which generally occur in 
clusters with marshy depressions between them. They are thought to 
have been formed at the glacier's edge, where steep cones of water-laid 
gravel became heaped up against the ice front, and, as the ice front 
melted, slumped down as irregular hillocks; or under the frontal ice 
by gravel-laden streams that flowed into depressions or holes; or in 
shallow basins of water on top of the ice, from which position the strati- 
fied gravels slumped into piles as the ice melted. 

Eskers are long, winding ridges of stratified gravel or sand that look 
like railroad embankments. They are only a few feet wide at the top, 
and range from 10 to 100 feet in height, but some of them may be traced 
for many miles in length. They are striking features of the land- 
scape in Scandinavia, Finland, Ireland, and the northeastern United 
States (Fig. 97). They are supposed to have been formed by deposition 
from streams that flowed in ice tunnels in and beneath stagnant ice. 
As the stream beds became built up by deposition of gravels, the streams 
enlarged their tunnels by melting the roofs. Thus there accumulated 
long, sinuous deposits of water-laid gravels enclosed by ice. When the 



ice melted, the gravels slid down to their angle of repose, to form eskers 
which trend in the general direction of ice movement. 

Fig. 97. The esker of Punkaharju, Puruvesi, Finland. In Scandinavia such a ridge 
is called an "ose" (plural "osar"). 

Lake Deposits. Remarkably banded clays have been formed in 
patches within the glaciated regions, notably in Scandinavia and North 
America. The Swedish geologists call them varved clays. In a fresh 
excavation it is evident that they are horizontal and that most of the 
individual layers are an inch or less in thickness. (See Part II, Fig. 
229.) But the observer is impressed chiefly with the striking uniformity 
in thickness of the separate bands and the sharp line of demarcation 
between layers. If he examines a band closely, he will see that slightly 
coarser sediment at the bottom grades upward into material so fine that 
it will not grit between the teeth, and that this alternation is repeated 
in the individual bands throughout the depth of the deposit. Such fine 
rhythmic banding surely could have been formed only in bodies of quiet 
standing water, such as lakes, and the repeated alternations of coarse 
and fine material suggest at once a seasonal deposition. And that, 
apparently, is the manner in which they were formed. The lakes 
were temporary features, ponded by morainal material or in part by 
stagnant ice, beyond the melting and retreating ice front. Into them 


flowed graded glacial streams carrying rock flour. The coarser material 
was deposited during spring and summer, when the maximum flow of 
water entered the lake; the finest material, at the top of each varve, 
represents the slow settling during fall and winter, while the lakes and 
streams were frozen over most of the time. Thus each layer, grading 
from coarser to finer material, represents one year's accumulation. 
Professor De Geer, in Sweden, realizing this, counted the annual layers 
and reached the conclusion that Stockholm was under ice 9000 years 
ago; that the ice sheet began to leave northern Germany 17,000 years 
ago, and southwest Sweden 12,000 years ago. He also found that it 
took about 5000 years for the ice to retreat to a point 270 miles north- 
west of Stockholm. Similar counting in Ontario by Antevs showed that 
it has been 13,500 years since the ice receded from north of Georgian Bay. 


The retreating ice cap had along its front places where the ground 
sloped toward It, and naturally In such sites water accumulated to form 
lakes. These disappeared when the impounding ice melted. Lakes of 
this origin, some large, some small, abounded along the ice fronts; 
the beaches, outlets, and deposits are still visible. One of the most 
noteworthy of these, known as Glacial Lake Agassiz y covered a large 
area in North Dakota, Minnesota, and southern Canada. It was 
700 miles long by 250 miles wide. The only remnants now left of it 
are the present Lake Winnipeg, Lake Manitoba, Lake of the Woods, and 
other smaller water bodies. Its extensive deposits constitute the fertile 
wheat lands of that region. Similarly, a series of lakes of changing out- 
line occupied the area of the Great Lakes and St. Lawrence Basin and 
extended across eastern Ontario. Their outlet was for part of the time 
through the Mississippi River, later through the Mohawk Valley to the 
Hudson River, and finally through the St. Lawrence. Abundant evi- 
dences of these old lakes are to be found in well-preserved beaches and 
lake deposits in the vicinity of the present Great Lakes. 


In the light of our knowledge of the behavior and results of existing 
glaciers, the features just described under " continental glaciation " 
lead inevitably to the conclusion that extensive ice caps once overspread 
great continental areas during the Glacial Period. The evidence is 
now so plain and convincing that it is difficult to realize that the idea 
was first proposed in 1837 by Louis Agassiz, and that a quarter of a 
century elapsed before the glacial theory was generally accepted. The 


older ideas that the debris was transported by the Deluge, or by floating 
icebergs moved by the Deluge, were dispelled slowly. 

If one were to travel widely over the glaciated lands, measure the 
directions of the glacial striae, roches moutonnees, eskers, drumlins and 
terminal moraines, and plot them in their proper place on a map, he 
would find that the glaciers did not originate at the poles. In North 
America there were three main centers, from which the ice spread out- 
ward in all directions. In Europe the main ice cap centered in Scan- 
dinavia and spread into Russia, Germany, Holland, and the North Sea. 
The thickness of this ice sheet is thought to have been 6000 to 7000 feet 
in Scandinavia and about 1500 feet in the Harz Mountains. The 
Labrador sheet was at least 6000 feet thick in northern New England, 
and may have been thicker farther north. 

At the same time that the great ice caps were spreading over the lands, 
mountain glaciation was also extensive; and while the ice caps of 
America and Europe disappeared, the mountain glaciers receded to their 
present-day smaller proportions. About four million square miles of 
America and two million square miles of Europe were glaciated, and 
many mountain areas in addition. There still remain another six mil- 
Lion square miles covered by ice in Greenland and Antarctica, so that 
during the Glacial Period about one-fifth of the land area of the globe 
was ice-ridden. 

Extended studies of the deposits of the Glacial Period show that there 
was not a simple advance and retreat of each ice cap. Glacial history 
is far more complicated. In North America there were at least four, 
and perhaps five, separate advances and retreats of the ice. These are 
recognized by later till sheets that cover earlier drift, or that overlie 
interglacial soils, and by other criteria. There was a widespread retreat 
of the ice in each interglacial epoch, although it is not certain that the 
ice disappeared entirely preceding each readvance. 

The recession of the ice caps left the lands greatly changed. Not 
only were they eroded, covered by deposits, and dotted by lakes, but the 
preglacial drainage systems were profoundly altered. New streams 
originated, old channels became buried, directions of flow became re- 
versed, and many waterfalls were formed. Deranged drainage on a 
large scale is evident in the Ohio River basin and in parts of New England. 

But what caused these great continental ice caps? The answer to 
this question has puzzled investigators ever since the fact of continental 
glaciation became firmly established, but as yet there is no generally 
accepted explanation. It is becoming clear, however, that they prob- 
ably did not owe their origin to a single cause but to a combination of 
causes. The various hypotheses advanced are discussed in Part II 


of this book. Here, it need only be mentioned that the most important 
factors in the origin of continental ice caps seem to be variations in 
amount of solar energy received from the sun which in turn affect 
the storminess on the Earth and the amount of heat stored in the oceans 
and changes in the shape and elevations of continents and mountain 


1. The Natural History of Ice and Snow, illustrated from the Alps; by A. E. H. 
Tutton. 319 pages. London, 1927. 

Excellent description and photographs of Alps glaciers; popularly written; 
good material on properties of ice. 

2. Ice Ages, Ancient and Recent; by A. P. Coleman. 296 pages. Macmillan, 
New York, 1926. 

Good description of continental glaciation, popularly written. Causes of glaeiation 

3. The Quaternary Ice Age; by W. B. Wright. 464 pages. Macmillan, London, 

Sound, comprehensive, finely written, well illustrated. 

4. Characteristics of Existing Glaciers; by W. H. Hobbs. 301 pages. Mac- 
millan, New York, 1911. 

Entertainingly written; many valuable data on present ice caps. 

5. Alaskan Glacier Studies; by R. S. Tarr and Lawrence Martin. 498 pages. 
Nat. Geog. Soc., Washington, 1913. 

A wealth of interesting data. 



Water on the surface of the Earth wears away the highlands and 
fills the depressions with sediment. It is a leveling agency that tends in 
the long course of time to make the land smoother. But in addition to 
the work of leveling at the surface, water performs other geological work 
of great importance, chiefly chemical, beneath the surface. This chap- 
ter discusses first the nature, position, and motions of the subsurface 
water, and the forces that control it, then the work that it performs. 

Source and History of Subsurface Water. A part of the sub- 
surface water ascends directly from the reservoirs of fluid rock within the 
Earth; another part is water that was trapped in the sedimentary strata 
at the time the sediment was deposited; but most of it is that part of 
the rainfall that sinks into the soil and into the bedrock below. This 
water, once it has penetrated into the ground, may find its way back to 
the surface as springs and join the run-off; it may be drawn to the sur- 
face by capillary attraction through the pores in the soil and be evapo- 
rated; it may be sucked up by plants and be evaporated through their 
leaves; it may find its way to the sea through underground channels 
without returning to the surface; it may be held for indefinite periods 
within the pores of the rocks; or it may form combinations with the 
molecules of certain minerals and so become fixed in the rocks. 

Porosity of Soils and Rocks. The porosity of a material is its prop- 
erty of containing open spaces or interstices and may be expressed as 
the proportion of the total volume of the material not occupied by solid 
matter. This property varies greatly in different materials but the 
existence of subsurface water depends upon the fact that all rocks and 
soils are more or less porous. 

It has been shown that a group of spheres of equal size placed to- 
gether so that each one touches all those that surround it leave unfilled 
25.95 per cent to 47.64 per cent of the total space that they occupy. 
The unfilled space is pore space and the difference in amount depends 
upon the manner in which the spheres are grouped and not at all upon 
their size. Ordinary sedimentary matter is not composed of perfect 
spheres but of irregularly shaped, more or less rounded grains of varying 




size. If these grains are well sorted so that those of one size are col- 
lected together to the exclusion of other sizes, the porosity of the material 
approaches and, because of the irregular shape of the grains, may even 
exceed the maximum for spherical grains. Poorly sorted sedimentary 
^materials have a much lower porosity because fine grains fill the inter- 
stices between the larger ones and reduce the amount of open space. 

The porosity of materials of the same general kind varies considerably 
but the following table gives a good idea of average amounts. These 
figures show that soils have the highest porosity in spite of the poor 
sorting of their constituents. This i explained by the poorly compacted 
condition of most soils. They are kept loose and aerated by worms, 
ants, burrowing animals, and vegetation as well as by plowing and agri- 
culture in general. 

Unconsolidated Material 

Consolidated Material 
(Sedimentary Bocks) 

Igneous and Metamorphic 

Sand and gravel 35 per cent 
Clay 45 
SoH 55 

Sandstone 15 per cent 
Shale 4 
Limestone 5 K 

Average porosity less 
than one per cent 

Igneous, metamorphic, and well consolidated sediment ary rocks 
commonly have a porosity much higher than that given because systems 
of fractures develop within them. Limestone may be rendered very 

Fig. 98. Diagrams to show relation of rock texture to porosity, a, well-assorted 
sedimentary deposit having high porosity; fe, poorly-assorted sedimentary deposit having 
low porosity; c, well-assorted sedimentary deposit made of pebbles which are themselves 
porous, giving the deposit as a whole very high porosity; d, well-assorted sedimentary 
deposit whose porosity has been diminished by deposition of mineral matter in the inter- 
stices; e, rock rendered porous by solution; /, rock rendered porous by fracturing. (U. S. 
Geol. Surv.) 

porous by the development of cavities due to solution. Figure 98 and 
the accompanying description illustrates the relation between rock 
texture and porosity. 

Aquifer. The term aquifer is applied to a layer or other body of 
loose sediment or consolidated rock that yields water in considerable 


Forces that Control Water in Rocks. Gravity and molecular at- 
traction are the two forces that control the movement of subsurface 
water. Gravity causes the water to percolate to lower levels or to move 
laterally down a grade. It also causes the water to issue from springs 
and flowing wells, forced out by the weight of the water behind. If all 
the interstices in rocks were large, gravity would be the only controlling 
force. But, since the actual interstices in many rocks are minute, the 
molecular attraction of the rock substance acts across the openings and 
holds the molecules of water firmly in place in spite of their tendency to 
move downward. This force is called adhesion. 

Water- Yielding Capacity of Rocks. Not all of the water contained 
in the pore space of a rock will be yielded to the natural drainage or to 
such artificial openings as wells. It has been shown that the size of 
the grain of a sediment does not affect its porosity but the last paragraph 
explains how that size does determine the percentage of the contained 
water that the sediment will yield. A porous and thoroughly saturated 
clay may be impervious to water at usual pressures because the water 
already in it is held firmly by the molecular attraction of the particles 
of clay. On the other hand a compact, granular rock such as a granite, 
with a few scattered fractures, is much less porous but yields the water 
that it does contain more readily. It is more pervious than the clay. 


The water in the ground fills all the pore space up to a certain definite 
level called the water table. The zone below the water table is known as 
the zone of saturation and the water contained in it is called ground water. 
The region between the water table and the surface of the ground is 
called the zone of aeration and the water in it suspended subsurface water 
(Fig. 99). The water table is not a level surface but rather a subdued 
replica of the land surface beneath which it lies. It is arched up under 
the hills, sinks a little below the valleys and intersects the surface in 
some low lying tracts to form lakes, swamps, or springs. 

The level at which the water table stands in any particular region 
depends on the rainfall and the topography. It rises or falls in response 
to wet or dry seasons and stands within a few feet of the surface in humid 
regions though it may be several hundred feet in depth in those that are 
perennially dry. If climatic conditions are the same the water table 
will stand farther below the hill tops in a rugged region than in one less 
rugged because the deeper valleys depress the water table near them 
and give the wkole a greater gradient. 

The Zone of Aeration. The thickness of the zone of aeration varies 
directly with the depth of the water table. It may be nonexistent or 



it may be several hundred feet thick, though' ordinarily it varies between 
a few feet and a few tens of feet. It is the zone in which suspended 
water, present in small quantities only, is held in place despite the 
downward pull of gravity. This zone is subdivided into three parts: 
(1) The soil water zone, the zone in which water is retained by the minute 
interstices of the soil and is thus available for plant growth. The 

Fig. 99. Block diagram to show the relation of the water table to the surface of the 
ground; the direction of movement of subsurface water (indicated by arrows); and the 
zones of aeration and of saturation. 

The upper part of the diagram shows the surface of the ground and the downward 
motion of the subsurface water as far as the water table. The lower part of the diagram 
shows the surface of the water table a subdued replica of the ground surface and the 
motions of the ground water in the zone of saturation. Notice that the ground water 
moves both down the slopes of the water table and down into the ground, and tends to 
converge at and below the main drainage channels of the region. 

The two parts of the diagram are separated in order to show the water table and the 
movement of the water along it clearly. The dotted lines A and .4' would coincide if the 
two blocks were in proper position. 

The subdivisions of the zone of aeration are given in the text. 

For the sake of simplification the region illustrated is regarded as underlain by a single 
stratum of uniform porosity. The difference in the density of the dots represents the 
different amounts of water contained. 

quality of a soil that allows it to hold water in this manner is its specific 
retention. A poorly sorted, compact soil has a high specific retention 
while a surface layer of material that yields water readily, such as pure 
sand, has very low retention. (2) The capillary fringe, the zone imme- 
diately above the water table that contains water drawn up by the 
capillary openings in the rock or soil. This zone is thick if the inter- 
stices are minute, as in clay or loam, and unimportant if they are 
coarse, as in gravel or sand. (3) The intermediate zone normally lies 
between the capillary fringe and the soil water but may be absent where 
the water table is near the surface. It is the zone in which a small 
amount of the water that percolates from the surface down to the zone 
of saturation is retained by molecular attraction. 



The Lower Limit of the Subsurface Water. Deep down within the 
crust of the Earth the weight of overlying matter exceeds the crushing 
strength of rocks. At such depths the interstices must c^ase to exist 
and consequently subsurface water is excluded. This theoretical depth 
is impossible to state accurately but it lies a number of miles beneath 
the surface. Actually ground water becomes very scarce after a few 
thousand feet of rock have been penetrated. Many mines are dry 
2000 or 3000 feet down except where deep fissures are encountered. 
Some aquifers have been encountered at depths as great as 6000 
feet, but as a rule wells drilled more than 2000 feet produce little 

Ground water must be regarded then as a great body of water filling 
the interstices in all soils, loosely consolidated sediments, and solid rocks 
between a sharply defined upper boundary called the water table and an 
indefinite lower boundary ranging from a small depth to many thousand 
feet below the surface. Lower than this even the most persistent fis- 
sures must be closed and dry. 

Perched Water Tables and Impervious Beds Lying Below the Water 
Table. The conditions outlined above are not always completely 

Fig. 100. Showing perched water tables caused by impervious beds interstratified 
with pervious beds, and the depression of the main water table by an impervious bed that 
extends down into the main zone of saturation; a and b are perched bodies of ground water; 
c is the main zone of saturation. (Adapted from Gregory.) 

fulfilled. In regions where the main water table lies at a considerable 
depth below the surface an impervious stratum in the zone of aeration 
may hold a local zone of saturation suspended far above the main one. 
This zone has its own water table called a perched water table (Fig. 100). 
There are also many impervious strata that extend far below the general 
level of the zone of saturation. A well that penetrated these strata 
would be dry even after it had passed some distance below the general 
level of the water table in the surrounding region. Many such imper- 
vious strata would cause the water table to be ill-defined. 




Motions of the Subsurface Water. Above the water table the move- 
ment of subsurface water is directly downward (Fig. 99). Below that 
level the ground water proper moves in the same general direction as 
the surface run-off but more slowly, because the friction caused by its 
passage through the interstices of the rock retards its motion. The 
continuous addition of parts of this slowly moving sea of ground water to 
streams and lakes and its seepage at the surface or concentration into 
springs is the cause of the continuous flow of streams. Were it not for 
the restraining influence of friction the subsurface flow T would be as rapid 
as that at the surface and every rain would mean a flood and every stream 
would run dry between storms. 

The more deeply the ground water penetrates beneath the surface in 
uniform material the more resistance it meets due to friction and the 
more slowly it moves until, at the greatest depths to w r hich it can go, it 
is stagnant, held motionless by the molecular force of the rocks. 

Hillside Springs. Wherever the surface of the land intercepts 
the water table, ground water emerges in some fashion. When the 


Spring or Seepage 

Fig. 101. Showing conditions favorable for wells, hillside springs, and a swamp or lake. 

water oozes out along the line of contact a seepage results, but if the cir- 
cumstances are such that water is concentrated in quantity sufficient 
to form a distinct current, an ordinary spring is formed. This is illus- 
trated in Fig. 101. Figure 102 shows the wall of a canyon that has cut 
the water table so that the water issues in a series of springs, 

Wells. An ordinary well is an opening dug or drilled into the ground 
to a depth sufficient to penetrate the saturated zone (Fig. 101). The 
water percolates into the opening from the saturated material around it 
but does not rise in the well above the level of the water table. 

Fissure Springs. When surface water enters and saturates a per- 
vious, inclined layer confined between two impervious layers the weight 
of the water in the aquifer generates hydrostatic pressure. If the aquifer 
is cut by a fissure that leads to the surface, the water will be driven up 
along the fissure and issue as a spring or a series of springs lined up along 



the fissure. Such springs are often called fissure springs. Figure 103 
illustrates the following necessary conditions: a catchment area where 
water may enter the aquifer; a pervious stratum to serve as aquifer 

Fig. 102. Thousand Springs, Snake River canyon, Idaho. The wall of the canyon 
cuts the water table and allows the ground water to issue as a series of springs. (U. S. 
Geol. Surv.) 

lying between two impervious strata; an inclination of the strata suffi- 
cient to cause the water in the pore space to exert pressure on that down 
slope from it; a fissure to cut off the normal course of the water and 
furnish a channel to the surface; a point at which the water may issue 

Fig. 103. Illustrating conditions favorable for springs, if fissures, such as/, are present. 

below the level at which it entered the aquifer. The difference in eleva- 
tion between these two levels is known as the hydraulic head which, 
together with the size of the pore space, determines the force with which 
the water will reach the surface. There is always a considerable loss 
of head due to friction. 


In such an arrangement as is postulated in Fig. 103, springs might be 
expected at various points along the line of intersection of the fissure 
with the surface, wherever suitable channels for the upward flow of the 
water exist. Such springs are usually very steady in their flow and are 
less affected by droughts than ordinary hillside springs. They are usu- 
ally cold, but the water may come in contact with heated rocks and issue 
as warm springs. This is probably the explanation of the warm springs 
that occur at various places in the Appalachians, as at Hot Springs, 
Virginia. The water sometimes dissolves unusual amounts of mineral 
matter and gives rise to mineral springs, as at Saratoga, New York; 
Carlsbad, Bohemia; Bath, England; Wiesbaden, Germany; and Yichy, 
France. These springs contain considerable though different amounts 
of sodium, chlorine, carbon dioxide, sulphate, and smaller quantities 
of calcium, magnesium, and many other elements. There are also 
springs high in silica, such as Hot Springs, Arkansas; Olette in the 
eastern Pyrenees; and the geyser waters of Yellowstone Park, Iceland, 
and New Zealand. It must be 
realized that, strictly speaking, 
all springs are mineral springs, 
since they contain mineral matter 
in solution. The term is indefi- 
nite but is applied in general to 
those springs that differ markedly 
from ordinary potable water, 

either in the quantity Of mineral Fig. 104. To illustrate entrance of water 
TYiQ-i-^r in Qnlntinn or in i*f^ ohar- ^^ Pervious rock layers, or strata. ACE, 

matter in solution or in its cnar beds . BDt pervious beds; RR, 

acter. course of river. 

The condition under which per- 

vious beds become filled with water is important, not only for fissure 
springs, but also for artesian wells, described below. It is illustrated 
in Fig. 104. The pervious layers BD become filled, not alone by the rain 
that falls on their exposed surfaces, and by the water that is shed upon 
them from the higher impervious slopes A and C } but from the river 
water that is concentrated from the watershed above. 

Artesian Wells. If under conditions similar to those described 
above, but in the absence of a fissure, a hole is bored down to the aquifer, 
the water will rise above the zone of saturation and produce an artesian 
well. The height to which the water rises above the zone of saturation 
depends upon the pressure, which in turn depends on the height of the 
water column, or " head " in the aquifer above its upper surface where 
the well is drilled; the size of the pore space of the aquifer, which de- 
termines the loss of head due to friction; and perhaps to some extent 



on the weight of the overlying formations which tends to compress the 
water-bearing stratum and force the water out. In many cases the 
water rises to the surface or above it. The alternating pervious and 
impervious beds may have the form of a basin (Fig. 105, A) but that is 
not a necessary condition. The arrangement illustrated in Fig. 105, B is 
just as suitable; the water rises through the artificial opening because 
the diameter of the well is greater than that of the openings that con- 

Fig. 105. Sections showing conditions favorable for artesian wells. Vertical scale 

A. Strata are folded so as to form a basin. 

B. Strata are inclined in one direction only. 

stitute the pore space of the aquifer and the water encounters less re- 
sistance than by pursuing its course underground. Figure 106 is a pho- 
tograph of an artesian well that shows the water spouting up as in a 

The conditions outlined above are possible only in sedimentary strata 
but the principle is applicable to other rocks as well. It is essential only 
that the water be confined, that sufficient head be developed, and that 
fractures or solution cavities act as passages for water. Such conditions 
are relatively rare and very local. They do not compare in extent with 
the widespread artesian basins developed in sedimentary strata. 

Artesian wells cannot be made simply by boring deeply unless the 
requisite geologic conditions are present. Deep wells bored into rock 
so as to intercept the water table are often called artesian wells but this 
is an incorrect use of the term; there is no difference in principle between 
wells of this kind and ordinary shallow dug wells. Some of the most 
important water-bearing formations in the United States which furnish 
artesian wells are the Dakota sandstone, which comes to the surface 
along the Rocky Mountains and Black Hills, and underlies large parts 
of North and South Dakota, Kansas, and Nebraska, and extends into 
Canada; the Saint Peter sandstone, which outcrops in central Wisconsin, 
and underlies much of Illinois, Indiana, Iowa, Ohio, Missouri, and Arkan- 



sas; and beds of sand that underlie the Atlantic Coastal Plain from Long 
Island to Texas. The conditions are generally unfavorable for artesian 
wells in the uplands of New England because the veneer of glacial till 
and lake and stream deposits is too thin and discontinuous to form 
artesian aquifers, and it overlies jointed and faulted metamorphic or 
igneous rocks in which the openings are too closely spaced or the in- 

Fig, 106, Artesian well, near Provo, Utah. (U. S. Geol. Surv.) 

dividual fractures are not sufficiently large to establish artesian cir- 
culation. In the lowland of central Connecticut numerous fractures 
prevent any great accumulation of water under pressure in otherwise 
well-situated sandstone strata. There are exceptions to this rule, 
however, and small local artesian basins are found scattered throughout 
the region. 

The depth to which wells must be bored before artesian water is 
attained is very great in some places. In Berlin, St. Louis, and Pitts- 
burgh the necessary depth is about 4000 feet, and depths of 1000 feet 
are not uncommon. Along the Atlantic coast, on the other hand, arte- 
sian wells are generally shallow from 100 to 300 feet. The volume 
of water may be very large; the great 12-inch well of St. Augustine, 
Florida, with a depth of 1400 feet, supplies 10,000,000 gallons a day. 


Where many wells are put down close together the drainage basins are 
likely to interfere with each other, and the withdrawal of water lowers 
the pressure to such an extent that the water will no longer rise above 
the level of the aquifer. 


Water underground is an important geologic agent. Its chief work 
is to take substances into solution, carry them elsewhere, and perhaps re- 
deposit them; this work is, therefore, largely chemical in its nature. 
Although this work may seem insignificant, the total results accomplished 
during geologic time have been enormous. A part of it has already been 
considered ; thus in the description of the decay of rocks and the forma- 
tion of soil it was shown that certain constituents such as the alkalies 
in the rock-forming feldspars go into solution and are removed; and 
that calcium carbonate, a common rock-making material, is dissolved 
and carried away by water containing carbon dioxide. These actions 
are accomplished by atmospheric water as it passes underground, and 
may thus be regarded as the first stages of the work of Subsurface water. 
Again its work was considered, when it was shown that rivers carry a 
large part of their burden in solution, and ultimately deliver this material, 
dissolved by subsurface water, to the sea. Finally the formation of 
salt lakes, and the deposits that occur in them, illustrate the work of 
solution, transportation, and deposition. 

These facts illustrate the general chemical work of water, partly on 
the surface and partly underground, but there are certain features that 
demand particular consideration. 

Solution. The solvent action of rain water passing into the soil 
and rocks is greatly increased by the substances that it carries with it, 
or that it may otherwise obtain. In its passage through the air it dis- 
solves carbon dioxide and oxygen, together with minute amounts of 
other materials, and is thus equipped for doing chemical work (Fig. 107). 
In passing through the humus and upper soil of humid regions it may 
absorb much more carbon dioxide as well as organic acids produced by 
the decomposition of organic matter. In many places, particularly in 
volcanic regions, volatile substances, especially carboy dioxide, are 
evolved from the depths, and may dissolve in the subsurface water, thus 
augmenting the quantity of chemical reagents present in it. In addition, 
as the water passes into deeper zones the pressure increases and it may 
come in contact with heated rocks and have its temperature raised, both 
of which changes greatly increase its chemical efficiency. The amount 
of gas, such as carbon dioxide, which water can hold in solution, is 



directly proportional to the pressure. The amount decreases as the 
temperature of the water rises but this is more than compensated by the 
increased chemical activity of hot water. 

Pure water dissolves many substances, but with its chemical efficiency 
heightened as just described, it attacks the minerals composing the 
rocks and soils with added energy. It takes some of them, such as 
gypsum (CaS0 4 .2 H 2 0), directly into solution. In other cases a chemical 
reaction takes place and new compounds are formed, some of which are 
soluble and are carried away, while the insoluble material remains. 

Fig. 107. Rock whose more soluble parts are being dissolved by the action of atmos- 
pheric waters. Wind aids the rain in removing the loosened material. Near Livingston, 
Mont. (U. S. Geol. Surv.) 

The decay of feldspar, as described under the formation of soil, is a 
good illustration of this process. The alkaline carbonates produced 
are leached out, whereas the insoluble clay remains. 

The material taken up and held in solution may pursue one or two 
courses, depending on what happens to the water that contains it. It 
may work down into the rocks and be deposited there, or it may emerge 
into the surface drainage and be carried into the ocean. 

The process by which the land surface is wasted by solution is known 
as chemical denudation, to distinguish it from the mechanical wear of 
ordinary erosion. The amount of material so removed each year is very- 
great. Dole and Stabler by assembling a large number of analyses of 


the waters of the Mississippi, which analyses give the average percentage 
of the salts it contains, have calculated that 108 metric tons of matter 
are removed each year in solution for every square mile that it drains. 

Results of Solution of Carbonate Rocks. Aside from the process 
of soil formation, in which solution plays an important part, the most 
obvious results of solvent action are seen in the effects it has upon rocks 
wholly, or partly, composed of carbonates. The most important rock- 
forming carbonates are those of calcium, magnesium, and ferrous iron; 
CaC0 3 , MgC0 3 , and FeC0 3 . The carbonate of calcium especially 
underlies vast stretches of land, as beds of limestone hundreds or even 
thousands of feet thick. Besides this, beds of sandstone commonly 
contain a cement of calcium carbonate that binds the grains of sand 
together. Since these carbonates, especially calcium carbonate, are 
attacked by water containing carbon dioxide, many such rock masses 
must be continually dissolving and wasting away. This is suggested 
by the fact that in those places where limestone, or calcareous sandstone, 
is the bedrock the water is always hard, i.e., contains lime in solution. 
Figure 107 illustrates the pitted or cavernous surface of a limestone ex- 
posed to the solvent action of water. The soluble material is carried 
off by the water and the insoluble residue breaks down and is blown away 
by the wind. 

Sinks and Caverns. In regions where limestone forms the bedrock 
the surface water works down through joints and fissures and enlarges 
them by solution. When the water reaches an insoluble layer, such as 
one of clay or shale, it is stopped in its descent and spreads laterally, 
finding its way through the rock fissures along the natural drainage 
slope. These fissures are also enlarged by solution until they become 
distinct water channels. As the latter enlarge they form caverns, 
while the holes or pipes, leading down to them from the surface above, 
are termed sinks (Fig. 108). Sinks are also formed by the collapse of 
the roof, weakened by solution, into the cavern beneath. 

Some of the individual chambers hollowed out in the rock are 100 
feet or more high, and several hundred feet broad. They are connected 
by intricate passages. The floor on the insoluble stratum may be quite 
level for long distances though it is broken through in many places when 
the water excavates new passages and chambers at a lower level (Fig. 
109). The final result may be several sets of such rooms and galleries, 
one above the other. Some well known natural bridges have resulted 
from partial collapse of cavern roofs. 

The limestone regions of the Middle West and the South are noted for 
their caverns, some of the best known being Mammoth Cave in Ken- 
tucky, 10 miles or more long, with 30 miles of winding passages; Wyan- 



dotte Cave In Indiana, Luray Caverns In Virginia, and the Carlsbad 
Cavern, New Mexico, In some places the rocks are honeycombed with 
passages and almost the entire drainage may pass underground. Large 

Fig. 108. Opening to sink in limestone beds; near Cambria, Wyo. (U. S. GeoL Surv.) 

rivers disappear from sight, and after a devious journey below come to 
the surface again in a different drainage area. They may give rise to 
huge springs, such as Silver Spring in Florida which has a flow so large 

Fig. 109. Illustrating the formation of caverns and sinks in limestones. A A, clay 
beds; BB, limestones. The arch is the remnant of the roof of "a former cave, forming a 
natural bridge. DD, sinks, leading to caverns below. (Modified from Shaler.) 

that the resulting stream is navigable for small steamers. Other under- 
ground streams may be forced up as great springs in the sea, not far from 

Deposition and Cementation. From what has been previously 
stated it is clear that there is an upper belt In the Earth's crust where 
mechanical and chemical changes and destruction are going on. It is 
known as the zone of weathering, and extends downward to the level of 


the water table. Material is being constantly leached out of this zone 
and carried down in solution into the ground water. This matter is 
either carried away by the drainage, or deposited in the pores and other 
cavities in the rocks. The lower limit to which this can proceed is un- 
certain. It appears to depend on several conditions and it probably 
varies in different places. Thus the intervention of impervious rock 
layers, or the charging of the rock pores with gas under pressure, would 
binder/or perhaps prevent, further downward movement in a given area. 
It is in this belt, whatever its thickness, that the sediments are being 
solidified and cemented by the silica, calcium carbonate (calcite), and 
other substances deposited in them, and it is, therefore, known as the 
zone of cementation. 

There are several reasons why much of the material dissolved by the 
ground water is redeposited by the same agency within the pores of the 
rocks. The most important of these are (1) Evaporation of water in 
open spaces such as caves. (2) Loss of dissolved gas such as C0 2 by 
warming or evaporation. (3) Cooling of the water. (4) Lowering of 
the pressure. (5) Chemical reaction between the solution and the rock 
through which it is percolating. 

Replacement. The deposition of material from solution in ground 
water does not always take place in openings already present. Fre- 
quently the water will dissolve a certain substance and leave behind 
an equal volume of another substance that it held in solution. Thus a 
buried tree trunk may be slowly converted into stone by the solution 
of its vegetable matter and the deposition of opaline silica (Si0 2 .n H 2 0). 
The fact that such replacements take place volume for volume is shown 
by the complete preservation of the fine texture of the original material 
regardless of the relative size of the chemical molecules involved. 

Concretions. The capacity of ground water to dissolve and re- 
deposit is nowhere better illustrated than by the formation of concretions. 
Concretions are regular, rounded, or strangely shaped nodules that occur 
in sedimentary strata. They are due to deposition from the ground 
water which has dissolved the substance from other parts of the same rock 
or brought in some foreign matter. They are explained in more detail 
in the chapter on sedimentary rocks. 

Stylolites. Irregular, sutured partings characterized by interlocking 
columns or teeth from a fraction of an inch to several inches in length 
commonly penetrate soluble rocks such as limestone, dolomite, or 
marble, essentially parallel to the bedding planes. These partings are 
known as stylolites. (Fig. 110). They are formed by unequal solution 
along the opposite surfaces of a fracture or a clay seam. They are 
localized by original irregularities of that surface which allow the weight 



of the overlying rock to press more at one point than at another and 
by the varying solubility of different parts of the rock adjacent to the 
fracture or seam. Once started, the rock opposite the end of each column 
is dissolved away more rapidly than elsewhere and the teeth penetrate 
further into the opposite bed and develop striae along their sides. The 
insoluble material finally collects as a clay cap at the end of each column. 

Fig. 110. Stylolites in Tennessee marble. (U. S. Geol. Surv.) 

Some stylolites develop in relatively insoluble rock, such as quartzite. 
Wherever they occur they indicate the removal of a considerable quan- 
tity of rock in solution. 

Relation of Subsurface Water to Ore Deposits. Certain deposits 
of lead and zinc ores seem to be due directly to the concentrating action 
of subsurface water that dissolves mineral matter over large areas and 
concentrates it in a relatively small zone. Many deposits of copper, 
zinc, gold, and silver are believed to be formed by heated waters ascend- 
ing from the cooling magmas below a portion of the total ground 
water whose quantity we have no means of estimating. Still other 
deposits mainly iron ores ^ ' owe their value to the removal of worth- 
less mineral matter in solution and the consequent increase in the per- 
centage of the iron minerals in a given area. The .value of many im- 
portant silver and copper veins is due to the concentration of the ore 
minerals from a large to a relatively small part of the vein by ground 



water percolating downward, thus raising the metal content within the 
narrow zone to a profitable amount. 

Deposits in Caves. The same process that forms caverns also tends 
to fill them up. For ; after they have been opened by solution, the 
surface waters seeping down through the rock layers which form the 
roofs dissolve more calcium carbonate, and deposit it in the caverns, 
producing stalactites, stalagmites, and columns. The manner of their 

Fig. 111. Stalactites, passing below into stalagmites, along a roof-crack. 
Cave, Indiana. (U. S. Nat. Mus.) 


formation is as follows: a drop of water, charged with calcium carbonate, 
leaking through to the roof hangs there -for a time. While at rest it 
evaporates a little, loses some carbon dioxide, and consequently deposits 
some calcium carbonate. Finally, as more water is added from above, 
it is forced to drop, and falling on the floor below it evaporates still 
more and leaves another deposit. Thus long, pendant, icicle-like in- 
crustations called stalactites grow downward, and- broader, dome-shaped 
masses called stalagmites rise up from the floor immediately below. 
Finally these may increase so that they unite and produce columns. 
These forms are most likely to develop along fissures in the roof of the 
cave (Fig. 111). 



In past times caves have served as refuges for primitive men who in- 
habited them, or as dens for wild animals. Because of this the bones of 
men and animals, stone implements, and other objects have accumulated 
in them and been sealed up, in the deposits of calcium carbonate on 
their floors. Relics of this kind, especially in certain parts of Europe, 
have revealed much concerning the life and degree of culture existing 
in prehistoric time. 

Deposits of Calcium Carbonate by Springs. The material m solu- 
tion which is not deposited in the rocks is carried away by the drainage. 
It sometimes happens that on its way to the sea it comes to the surface 

Fig. 112. One of the terrace formations of the Mammoth Hotsprings, Yellowstone 

Park, Wyo. 

again and is temporarily deposited. Springs that deposit calcium car- 
bonate furnish a good illustration of this. Many springs, especially 
deep or fissure springs, contain much carbon dioxide gas, under con- 
siderable pressure. When this water passes through beds of limestone 
on its upward journey large quantities of calcium carbonate are taken 
into solution. On arriving at the surface, partly because of evaporation 
and partly because of loss of gas through the relief of pressure, the cal- 
cium carbonate is deposited and built up into mounds and terraces, 
some of which are very beautiful. They are illustrated in the basins 
and terraces of the Mammoth Hotsprings in the Yellowstone Park 


(Fig. 112). Other examples of such springs are found in Virginia, Color- 
ado, Banff in Alberta, Carlsbad in Bohemia, Tuscany, and in many 
other places. Some spring waters contain other mineral substances 
together with calcium carbonate, or even to its entire exclusion. Many 
of these springs are used medicinally, as at Saratoga and other health 

In some springs that come from great depth the issuing water is warm, 
or even hot. This is likely to be the case when the springs occur in 
regions of active or recently extinct volcanoes such as that in which the 
Mammoth Hotsprings are situated. In such warm waters the deposit 
of calcium carbonate may be much increased by the action of primitive 
forms of vegetable life, the algae, which secrete this substance from the 
water. In many places the deposition takes place so rapidly that ar- 
ticles suspended in the water become covered in a few days with a coat- 
ing of the mineral. It is probable that the warmth and chemical activity 
of the waters of some springs, particularly hot springs in volcanic regions, 
are greatly increased by gases and vapors that rise from molten magma 
or masses of hot rock below. Since water vapor is believed to form the 
largest part of these discharged gases the volume of a spring may be 
increased in this way. 

Nature of Calcium Carbonate Deposits ; Travertine, Tufa. The 
character of the material formed when calcium carbonate is deposited 
from solution depends on circumstances, especially on the rate of depo- 
sition. When it is produced by slow evaporation, as in the case of 
stalactites in caves, it is a hard, compact, more or less crystalline sub- 
stance. Travertine, from the old Roman name of a town (Tivoli) in 
Italy where an extensive formation of the substance exists, is a general 
name for such deposits. " Mexican onyx " or " onyx marble " is a 
travertine with banded structure brought out by varied tinting from 
metallic oxides. When calcium carbonate is formed rapidly from springs 
the travertine may be porous or loose, or it may coat vegetation and be 
spongy or mosslike. Such less compact varieties are commonly called 
calcareous tufa, or sometimes calcareous sinter. Great deposits of this 
material are found around the shores of dried-up alkaline lakes, such as 
Pyramid Lake in Nevada where it encrusts the rocks of the enclosing 

It should be clearly borne in mind that these deposits are not original 
formations of calcium carbonate, in the sense in which we think of that 
word in connection with limestone; they represent in large part previ- 
ously existent calcium carbonate, such as limestone, or chalk, which has 
gone into solution, been transferred to another place, and redeposited. 
They exhibit a temporary stoppage of the material on its way to the sea, 


for it is the fate of all deposits of carbonates, if exposed to atmospheric 
agencies, to be dissolved and eventually taken to the ocean. 

Other Deposits by Springs: Iron Oxides, Silica, etc. Substances 
other than travertine may be deposited when underground waters issue 
at the surface. One of these is the hydrated oxide of iron, or, under 
certain circumstances, iron carbonate. This is a matter of importance 
because some extensive beds of iron ore have been formed in this way. 
Under certain conditions silica, sulphur, and gypsum may be de- 
posited, but since agencies other than those which have thus far been 
described are concerned in the process this matter will be discussed 

Alkaline Deposits. The soluble substances that are formed by the 
decay of the rocks in humid regions are quickly washed out of the soil 
and are carried by the drainage to the sea. In semiarid and desert re- 
gions where the rainfall is scanty there is not sufficient water to perform 
this function and the salts remain in the soil. At times of rainfall they 
go into solution, and during the subsequent periods of drought, when the 
water draws to the surface, they are left behind by its evaporation and 
form the white incrustation on the soil known as alkali. This is a com- 
mon feature in many parts of the western United States. 

The common salts that compose alkali are sodium sulphate, sodium 
chloride, and sodium carbonate, Na 2 S0 4 , NaCl, and Na2C0 3 . The 
name alkali is due to the alkaline reaction and taste of the latter. Mag- 
nesium sulphate, MgS0 4 , and calcium sulphate, or gypsum, CaS0 4 .2 H 2 0, 
are often present as well. These salts are not always furnished directly 
by rock decay; they may have been originally present in beds of sedi- 
ments laid down in the sea. Their concentration in such arid regions, 
with inland drainages, gives rise to salt and alkaline lakes. The irriga- 
tion of alkali lands, especially if water is too freely or carelessly used, 
may bring the salts to the surface in such quantities as to injure, or 
even ruin the land for agriculture. 

Mechanical Work of Water Underground ; Landslides. Subsurface 
water is of little importance as a mechanical agent. It is conceivable 
that streams running hi subterranean channels may erode and transport, 
but the circumstances that would permit this are exceptional. A mgre 
important mechanical function of subsurface water is its aid in causing 
landslides, both by helping to overcome the friction of masses of rock, 
soil, and debris lying on steep slopes, and by adding weight to the mass. 
The saturated masses of soil and rock act like a semifluid substance and, 
once started from their insecure foundations at times of heavy rainfall 
or when loosened by earthquake shock, rush down into the valleys caus- 
ing great damage and considerable changes in the topography. In 


high mountains such landslides may precipitate huge trains of broken 
rock, or talus, down the valley sides, giving rise to rock streams. 


1. The Occurrence of Ground Water in the United States, with a Discussion 
of Principles; by 0. E. Meinzer. 321 pages. U. S. Geological Survey, Water- 
Supply Paper 489, 1923. 

Accurate and authoritative discussion of principles, the kinds of rocks and their 
water-bearing properties, the influence of the rock structure on ground water, and the 
water-bearing formations of the United States. 

2. Domestic Water Supplies for the Farm; by M. L. Fuller. 180 pages. Wiley 
& Sons, New York, 1912. 

Good, nontechnical account of the problem of water supply with special emphasis 
on locating and constructing wells. Brief general treatment of the principles of 
occurrence of subsurface water. 

3. Underground Water Resources of Connecticut with a Study of the Occurrence 
of Water in Crystalline Rocks; by E. E. Ellis. 200 pages. U. S. Geological Survey, 
Water-Supply Paper 232, 1909. 

Good discussion of ground water in fissured crystalline rocks. 


Inland bodies of standing water are called lakes; expanded portions 
of rivers are sometimes also referred to as lakes. Small water bodies, 
particularly where shallow and filled with aquatic plants, are known as 
ponds, but there is no fixed usage as to these terms. From small ponds 
there exists every gradation up to Lake Superior, the world's largest 
freshwater lake, and the Caspian Sea, which though salt is the largest 
inclosed body of water. Thus, though most lakes contain fresh water, 
many are salty. The great majority of lakes stand above sea level, 
but some, including all coastal lagoons, are at sea level; a few, like the 
Salton Sea and the Dead Sea, are even below. Lakes occur in all parts 
of the world but since a great proportion of them are a direct result of 
glaciation, there are more lakes in high latitudes and in high altitudes 
than elsewhere. There would be no lakes if the surface of the lands 
were everywhere graded, for then drainage to the sea would be perfect. 
Where drainage is obstructed and the surface is not at grade lakes occur. 
Since all the streams are at work destrO3-ing obstructions and filling 
depressions in the ceaseless effort to bring the lands to grade, it follows 
that lakes are necessarily ephemeral and that all must sooner or later 


Warping and Faulting. Lake basins formed by broad downwarping 
of the rocks have existed at many times during the Earth's history and 
most of them probably have been large. We can be sure of this because 
their shore and bottom deposits are preserved long after the waters that 
stood above them have been drained away. 

Great lake-bearing depressions caused by fracturing and downfaulting 
of the crust are more common than those caused by folding. The 
4000-mile chain of valleys and lakes that includes Jordan River, the 
Dead Sea, the Red Sea, the upper Nile, and the African lakes, such as 
Tanganyika and Nyassa, was formed by the sinking down of narrow 
blocks of the crust, between high steep walls. This great " rift valley " 
contains more than 30 lakes, several of which are notably large and deep. 
Lake Tanganyika is about 5100 feet deep, and since its surface is only 
2500 feet above sea level its bottom is 2600 feet below. The Flatten 



See of Hungary, 50 miles long, the Warner Lakes of Oregon, and some 
of the larger lakes of southern Sweden likewise owe their basins to fault- 
ing. In certain cases the formation of such structural basins has oc- 
curred within human history in connection with earthquakes. In 1811 
an earthquake shook the lower Mississippi Valley and caused such 
changes in the surface that several new lakes came into existence in the 
Tennessee portion of the flood plain. 

Crater Lakes. One of the most remarkable natural features in 
North America is Crater Lake in southwestern Oregon, occupying a 
crater more than 5 miles in diameter in the summit of the Cascade 
Range. The arrangement of the rocks which form the crater shows 
that this was once a high volcanic mountain, but that the top collapsed 
and sank away 3 leaving a great caldera (see p. 253 and Fig. 176). The 
depression thus formed filled with rain water, forming a lake 2000 feet 
deep, without tributaries and entirely dependent on rainfall. The 
water escapes by evaporation and by seepage into the rocks of the crater 
rim, reappearing in part as springs at lower levels. Crater lakes are 
found in most volcanic regions. Notable among them are Lago de 
Bolsena and others in the great volcanic " campagna " surrounding 
Rome, Lac Pavin in the volcanic plateau of central France, and the 
" maaren " of the Eifel in northwestern Germany. 

Natural Dams across Valleys. Natural dams may be thrown across 
valleys in many ways. Chippewa River, which enters the Mississippi 
about 60 miles below St. Paul, has deposited sufficient debris athwart 
the course of the larger stream to restrict flow and the Mississippi has 
in consequence been ponded for more than 20 miles upstream. Land- 
slides occasionally interrupt the flow of streams, converting them into 
lakes. Several such lakes are known in the region of the Alps, and a 
new one was formed at the base of the Gros Ventre Range, in western 
Wyoming, in 1925. Lava flows occurring in regions where streams 
have already cut narrow valleys are likely to dam the valleys and form 
lake basins. The lava barriers are eroded with difficulty because made 
of solid rock. Lac d'Aydat in central France is of this type, as are 
several of the lakes surrounding the volcanic cones of Mt. Hood and 
Mt. St. Helens in Oregon and Washington. Again, where valleys are 
drowned and the force of their stream currents is dissipated by coastal 
submergence, it is easy for waves and currents to build bars and barriers 
across their mouths and so convert them into coastal lagoons. Long 
coalescent chains of these lagoons are found along the Atlantic coast 
of the United States from Sandy Hook to Florida. 

Glacial Lakes. The majority of existing lakes are the direct result 
of glaciation. This is well shown by the fact that most of the lakes of 


North America are concentrated within the northern half of the con- 
tinent, precisely within the area which has recently been glaciated. 
Some of these lakes occupy depressions scoured or plucked out of the 
bedrock (cirque lakes are among the most common); others, mere 
ponds, occupy kettle holes, and still others are ponded back of dumps of 
glacial debris. Most are small, but some large lake basins as well are 
of glacial origin. The southern end of Lake Michigan owes its outline 
to a great crescentic ridge of morainal material behind which it was 
dammed; while many of the large lakes of the Alps (Lucerne and Con- 
stance on the north and Maggiore, Lugano, Corao, and Garda on the 
south) are held in by moraines at the mouths of deep glaciated mountain 

Lakes on Flood Plains. Flood plains of streams, aside from the 
stream channels themselves, are areas of conspicuously poor drainage. 
Shallow lakes are numerous in cut-off and abandoned meanders- (oxbow 

Fig. 113, A temporary lake in a semiarid region. Wyoming. (U. S. Geol. Surv.) 

lakes), in abandoned temporary channels within single large meander 
loops, and in the depressions formed between natural levees and the 
outer edges of flood plains. Lake Maurepas above New Orleans is a 
large lake of the latter type! 

Limestone Sinks. Large areas in southwestern Kentucky and in 
southern Indiana are remarkable in that they are unglaciated uplands 
dotted with thousands of tiny lakes and ponds. These lakes occupy 
limestone sinks and are common also in parts of central Florida, Yucatan, 
Yugo-Slavia, and other limestone areas of the world. Sinks can con- 
tain water only (1) if their bottoms are below the water table or (2) if 
the outlets are clogged with insoluble clay left after solution of the 


caused by the Wind. Bare rock surfaces in arid regions 
such as the high plateaus of northern Arizona and Utah not infrequently 
contain shallow depressions hollowed out by wind-driven sand. These 
depressions sometimes contain intermittent lakes. The hollows be- 
tween live dunes also contain water in some cases, the water being pre- 
vented from sinking down through the sand by layers of decaying plant 
matter. Several large lakes of this kind are found among the dunes 
at the south end of Lake Michigan, and others are dotted about the 


Inherited Depressions. The sea floor contains many irregularities 
and depressions, and therefore an area newly uplifted from beneath the 
sea often contains such depressions. Most of the Florida lakes which 
do not occupy sinks are of this sort, One of the largest is Lake Apopka 
near Kissimmee. 

Playa Lakes. The broad basins between desert mountain ranges 
are sometimes filled to a depth of a few feet after the sudden and violent 

Fig. 114. Playa lake in a desert valley near Ludwig, Nevada. The lake is rapidly 

disappearing through evaporation. (Knopf.) 

rainfalls characteristic of such regions. Ephemeral or playa lakes 
are thus formed which 'disappear a few days or at most a few weeks 
after the rain has occurred. 

Relic Lakes. Imbedded in the sands and clays of the basin of 
Lake Champlain, geologists find the shells of marine molluscs and the 
bones of seals and whales. This is doubly remarkable in that the fossil- 
bearing deposits now stand 440 feet above the sea. The same fossil 
remains are found also in the St. Lawrence valley as far west as Lake 
Ontario. From these facts it is evident that the St. Lawrence and 
Champlain valleys were occupied in comparatively recent times by an 


arm of the sea which has since been forced to withdraw because of uplift 
of the land. Such lakes as Champlain, relics of former seas, are called 
relic lakes. The Caspian Sea is the largest example of this type. 


Lakes are constantly coming into being through the agency of these 
basin-forming processes, and as constantly those in existence are being 
destroyed. The broad playa lake in a desert basin is the most ephemeral 
of all : the waters begin to be sucked up by evaporation as soon as they 
come to rest. But its containing basin is not destroyed; and with the 
next cloudburst, the lake reappears as before. In moist regions, how- 
ever, evaporation goes on much more slowly, and at the same time the 
inflow of water into lakes is greater. Hence in these regions, lakes do 
not dry up. But here other forces are at work, which although they 
operate much more slowly, are far more dangerous to the existence of 
lakes because they destroy the basins which contain them. " Rivers 
are the mortal enemies of lakes" (Salisbury). The case of Lake 
Geneva and the river Rhone illustrates this striking statement. The 
lake of Geneva occupies a deep mountain valley and is at present about 
40 miles long. It was originally 7 miles longer, but each year the turbid 
Rhone, which enters the lake from the east, brings down great quan- 
tities of fine sediment from the glaciers at its source and dumps this 
material in the quiet water, forming a great delta which fills the upper 
end of the lake basin from side to side. The water is nearly a thousand 
feet deep and the delta is correspondingly thick, but its bank-like front 
is creeping farther westward each year, and in time it will destroy the 
lake. While this is going on the water which spills from the surface 
of the lake at its lower end is cutting down its channel and is thus lower- 
ing the lake outlet. The water, as it passes under the many bridges 
of the city of Geneva, is beautifully blue and clear. All of the sediment 
from the upper Rhone has been added to the delta or has dropped to 
the bottom during the slow passage of currents down the lake. The 
lake has thus acted as a great settling-basin depriving the lower Rhone 
of tools with which to cut. Hence downcutting of the outlet is retarded, 
but it cannot be stopped. If these processes continue, the western 
rim of f the lake basin will be lowered, and the water level will corre- 
spondingly drop, until the lake will be destroyed by this process if it 
has not already been filled up by the encroaching delta. 

Downcutting of the rim of a large lake basin is more forcibly illus- 
trated by Niagara Falls at the outlet of Lake Erie. The water at the 
brink of the falls is lowering its channel in a bed of limestone. At the 


same time the falls are retreating upstream at a rate which averages 
about 5 feet a year. When the falls have migrated up the river to a 
point opposite Buffalo, Lake Erie will have been largely drained. It 
not infrequently occurs that a lake may acquire a wholly new outlet 
through being tapped by a stream. An example seen by thousands of 
people every- year is found in Yellowstone Lake near the center of 
Yellowstone Park. In former times this lake was larger than it is now 
and was drained at a point on its southwest rim by a stream leading 
through Snake River to the Pacific. The present Yellowstone River, 
flowing north, was then a small stream engaged in rapidly lengthening 
its valley headward on a steep slope. The head of this young valley 
eventually reached the lake, tapped and partially drained it, and di- 
verted its waters into their present course through the Missouri to the 

In the South Park of Colorado, in the district west of Pikes Peak, 
a stream is flowing through a valley whose sides are made of layers of 
tightly packed volcanic ash. Examination shows that these layers 
were deposited in a narrow lake, and that they came from an active 
volcano close by. The delicate ash, sifting down through the lake 
waters, carefully protected and perfectly preserved the remains of 
plants and animals that lived in and around the lake. The lake basin 
was gradually filled with ash and lava and was converted into a stream 
valley; and from the exposed layers of ash there have been taken more 
than 1000 species of insects, 250 species of plants (including numerous 
trees), fishes, and birds, all representative of the life of the time during 
which the lake existed. The filled-up lake, known as Lake Florissant, 
has thus become a storehouse of great scientific value. 

It is thus apparent that all lakes will sooner or later be destroyed by 
(1) sedimentation in their basins or (2) drainage at their outlets, or both. 
To the former factor must be added the contributions made by wind- 
blown material, the accumulated bones and shells of lake-dwelling 
animals, and more important still, the remains of aquatic plants. 
(See pages 163, 164.) 


Lakes in Humid Regions. A depression is formed in a certain 
locality. Will it contain a lake and if so what kind of a lake? In 
general the answer depends on rainfall and hence ultimately on climate. 
Each of the lakes with which we are familiar in eastern North America 
and in western Europe has streams or springs flowing into it, and an 
outlet or spillway determined by the lowest point in the rim of the 
containing basin. Since these regions are humid the rate of evaporation 



is relatively slow, and the constant inflow and outflow not only prevent 
the lake surface from fluctuating greath r but also keep the water fresh. 
In such lakes plant life is usually abundant, and the accumulations of 
decaying vegetable matter greatly help to fill up the lake basins, which 
are likely gradually to become ponds and swamps before they dry up 

Lakes in Arid Regions. In the arid and semiarid regions of the 
world, where rainfall is slight and evaporation great, lakes rise and fall 
seasonally, and many dry up and disappear for months at a time. 

Fig. 115. Alkaline salt lake near Parma, Colo. (U. S. Geol. Sun*.) 

In fact the playa basins in the deserts which lie betw r een the scattered 
ranges of Nevada, Utah, Arizona, New Mexico, and Sonora contain 
water for the most part only after sudden and infrequent rains. Here 
as well as in the desert of Gobi in central Asia, the great deserts of 
western Australia, the intennontane basins of the Andes, and many 
other arid regions, evaporation is so rapid and continuous, the soil so 
porous, and the ground water table so far below the surface, that the 
streams which start seaward, fed by the rains and snows of high moun- 
tains, disappear long before they reach their destination. Some of 
them dwindle away, spreading out their sediments in great barren fiats 



covered with incrustations of salt and alkali which in wet seasons become 
shallow lakes and salt marshes. Other streams, where the water supply 
is greater, end in lakes. Wherever evaporation prevents the water from 
overflowing the rims of these interior basins, the lakes either are salt 
or are fast turning salt. Every river carries dissolved salt in quantities 
too small to be detected by taste. Slowly but steadily this salt is 
added to the total already in the water of the lake. It cannot escape 
by overflow and it cannot escape by evaporation. So year by year the 
proportion of salt to water increases until not only can it be tasted but 

Fig, 116. Islands of calcareous tufa in Pyramid Lake, Nevada. (U. S. Geol. Surv.) 

the water becomes undrlnkabie. Here appears the great menace of 
some deserts. The lonely traveler or prospector whose water supply is 
exhausted, half crazed by heat and thirst, reaches the lake he has seen 
in the distance. And if the lake is really a lake and not a mirage, he 
tastes and finds what is worse than no water water all but saturated 
with salt. The Great Salt Lake in Utah has a salinity of 18 per cent 
compared with less than 3| per cent for the ocean; its waters are so 
dense that bathers at the beach near Salt Lake City cannot sink, but 
float buoyantly upon the surface. As a swimmer wades back to the 
teach the water evaporates from his body, leaving it encrusted with 
tiny crystals of glistening salt. In the same manner slow shrinkage of 
the lake is leaving great dry beds of white salt to mark its former gently 



sloping shores. The chief salts are common salt (sodium chloride) 

and sodium sulphate. Calcium carbonate is also present and is com- 
monly deposited as granules on the bottom and shores. In some lakes 

Fig. 117. Salt deposits on the floor of Death Valley, Cal. The salt in the foreground 
Is etched into pinnacles by weathering processes, and is covered with wind-blown dust. 
Salt formed recently is white. The valley floor is here 250 feet below sea level. The 
car in the middle distance gives the scale. (Longwell.) 

this substance is deposited in great spongy, mosslike masses known as 

calcareous tufa. By covering large rocks with thick incrustations, this 
material produces many curious and striking forms (Fig. 116). 

Fig. 118. Former shore lines and wave-cut terraces of the ancient Lake Bonnevilie. 

Extinct Lakes. Observant travelers in Utah and Nevada have 
had their curiosity aroused by frequent sights of parallel rows of cliffs 
and great flat-topped terraces forming huge flights of steps up the moun- 



tain sides to levels as high as 1000 feet above the valley floors. When 
examined in detail, these " steps '' prove to be wave-cut cliffs and wave- 
built terraces, bars, and beaches, as well as deltas high and dry. These 
features are arranged in sets at different levels, and when traced along 
the various mountain ranges of the region, the highest set is found to 

mark the shoreline of a lake 
which must have had an 
area of 20,000 square miles 
(nearly as great as that of 
Lake Michigan) and a 
maximum depth of more 
than 1000 feet. The water 
must have been 850 feet 
deep at one time above 
the site of the Mormon 
Temple in Salt Lake City. 
The outlet was to the 
north, through Red Rock 
Pass to the Snake River, 
and thence through the 
Columbia to the Pacific. 
Since there are several sets 
of shorelines at different 
levels, we must conclude 
that the lake dwindled by 
stages from its former gi- 
gantic size to that of the 
Great Salt Lake, which 
represents the deepest pool 
in the bottom of the basin 
of its predecessor. Simi- 
larly, Pyramid Lake and its 
neighborsinNevadaare the 
residual pools of another 
huge water body, which rivaled in size the present Lake Erie. The Utah 
like has been named Bonneville and the Nevada lake Lahontan. The 
fact that some of the terraces of Lake Bonneville are associated with 
glacial deposits indicates that the lake had its origin in the Glacial 
Period. The presence of the ice must have made the climate much more 
humid than now, because since the ice retreated the outlets have been 
slowly cut down and at the same time the lakes have been gradually 
drying up. Salt deposits interbedded with the lake clays, all now high 

Fig. 119. Map of the former Lake Bonneville; ruled 
areas show present water-bodies. 



and dry, bear witness to the evaporation which has been going on during 
the time since the Glacial Period. Such deposits, so clearly related to 
the lakes, help us to understand that climates slowly change; and when 
we find beds of salt in the rocks of central Michigan, a humid region, we 
can be fairly certain that the climate of that region was dry at the time 
the salt beds were being deposited (Figs. 118, 119). 

The Salton Sea. The mam line of the Southern Pacific Railroad 
eastward from Los Angeles has the curious distinction of running for 
more than 60 miles through an area that lies below the level of the sea. 

Fig. 120. Delta of the Colorado River and the Salton Basin. Dashed lines show sea 
level. Longest dimension of block about 200 miles. 

The area is the Salton Basin. It is a basin without outlet, and its 
bottom is 273 feet below sea level. In its center is a shallow salt lake 
known as the Salton Sea (Fig. 120). In the recent geologic past the 
floor of this basin sank, and would have filled with sea water; but 
Colorado River, heavy with silt, discharged into the Gulf near what is 
now the southern end of the Imperial Valley and built up a delta and 
fan so large that they formed an effective dam at the head of the Gulf. 
Like other deltas this one was trenched by shifting distributary chan- 
nels of the river, which discharged water sometimes into the shortened 
Gulf and occasionally into the isolated basin, filling it with water. 


When dry, the basin was kept virtually a desert under the prevailing 
arid climate. The last natural discharge into the basin occurred per- 
haps 300 to 1000 years ago. In 1900 the trapped water had dwindled 
to a small salt lake. At this time the fertility of the alluvial soil was 
realized, the basin area was settled, and an irrigation canal was dug 
from Colorado River to the bottom of the basin. Since the floor of 
the basin was well below the baselevel of the river, the gradient of the 
canal was greater than the gradient of the river between its mouth and 
the head of the canal. The flow of water into the canal was inadequately 
controlled by head works designed to be protective. In 1905 the Colo- 
rado rose in flood, overtopped the headworks, poured into the canal, 
and following the new steep gradient, cut a great trench, in some places 
SO feet deep, swelled the Salton Sea to many times its former size and 
depth, and flooded the railroad right of way for more than 40 miles. 
The inundation of the basin resisted all efforts at permanent control 
for more than 18 months; but it was finally mastered. The railroad 
tracks were shifted from 200 feet below sea level to 150 feet below sea 
level as a precaution against possible later floods. The Salton Sea 
began to dwindle as soon as the abnormal supply of water was cut off, 
but now the waste water from irrigation keeps the level nearly constant. 


Large lakes are effective in modifying local climates by increasing 
atmospheric humidity and by cooling the air in summer and warming 
it in winter* Lakes of all sizes are very important as regulators of stream 
flow, acting as storage reservoirs and minimizing the height of floods 
in lower regions to which they are tributary. This fact has been recog- 
nized by the Egyptian Government in its project to increase artificially 
the size of Lake Tana in Abyssinia, one of the important sources of the 
Nile, in order to increase its effectiveness as a regulator of water supply 
in the Lower Nile Valley, an area of great economic importance. Simi- 
larly, suggestions have been made in the United States that a series of 
artificial lakes be constructed in the Mississippi drainage basin in the 
attempt to control near their sources such floods as caused the disaster 
of 1927. All lakes likewise act as settling basins for river sediment. 
Most of the streams tributary to Lake Erie are well loaded with mud, 
but the clarity of the water which pours out of the lake and over Niagara 
Falls bears testimony to the amount of material which is constantly 
being dropped upon the lake bottom. 




Swamps are areas of saturated ground. The majority of swamps 
represent a stage intermediate between lakes or ponds and dry land. 
Most lakes will in time become swamps, and many shallow basins alter- 
nately contain swamps and lakes according to the season. Swamps are 
commonly found in three types of regions, but these regions by no means 
exhaust the possibilities. (1) Swamps are both numerous and large 
on nearly level coastal plains which are former sea floors slightly uplifted. 

Fig. 121. Destruction of a small lake by formation of peat. The accumulating layer 
of peat (solid black) is fringed by aquatic vegetation consisting of water weeds and pond 
lilies while directly above it are encroaching semi-aquatic plants, mosses, and bushes. 

Such swamps are distinguished from salt marshes and are almost con- 
tinuous along the South Atlantic and Gulf coasts of the United States, 
chief among them being the Dismal Swamp in Virginia and North 
Carolina, and the Everglades in southern Florida. Some of these 
swamps may be old lagoons, uplifted together with the offshore bars 
by which they were shut off from the sea. (2) Flood plains and deltas 
with their basins formed by old channels and by natural levees contain 
great areas of swamp land. (3) Broad glaciated areas such as the 
greater part of the Great Lakes region of the United States, northeastern 
Canada, and the Baltic Plain of northern Germany are dotted with 
swamps, most of them small. 



Formation of Peat. In humid regions, the shores of small lakes 
ai>! protected coves of large lakes support an abundance of aquatic 
\vpnaflon such a^ pond lilies, water weeds, and rushes. As these 
pl::iit< I>, their substance decays In the water by slow oxidation, 
Wiik-h Is caused largely by the activities of bacteria. During the 
of the bacteria, waste products, antiseptic in their action, 

Fig. 122. Dismal Swamp, Va. The projections from the cypress roots serve to give 

them air; they extend downward into the mud and help to anchor the tree in the semi- 
liquid inass of the bog. (U. S. Gool. Surv.j 

are excreted; hence when this waste matter reaches a certain concen- 
tration in the lake water the bacteria can no longer exist, further decay 
is prevented, and the partially decomposed matter is preserved. This 
matter is brownish or blackish, has a high carbon content, and is known 
as peat. 

As peat is gradually built up along the shore of a lake, newer genera- 
tions of aquatic plants advance toward the center of the lake, and other 
types of vegetation such as mosses encroach over the peaty area which 
was formerly water. Thus the lake, surrounded by concentric belts 
of different kinds of plant s ? gradually decreases in size until it is ob- 
literated, and a bog floored with a thick accumulation of peat takes its 
place (Fig. 121). 


Economic Aspects of Peat. In many countries, especially in Europe, 
peat is cut from the bogs, dried, and used as a domestic fuel. In 
America peat has hitherto been little used because of an abundance of 
coal and wood. The peat resources in bog lands within the United 
States, however, are enormous, and will probably constitute a valu- 
able source of fuel in the future. 

Peat is of special interest to the geologist in that it represents the first 
stage in the transformation of vegetable matter into coal. All grada- 
tions may be observed from peat through lignite, bituminous (" soft ") 
coal, and anthracite (" hard ") coal. Thus if present peat bogs ? the 
best of which lie in temperate and cold humid regions, were left un- 
touched, many of them in the course of time would be covered with 
sediment, the remaining necessary changes would take place, and 
the result would be the formation of coal interbedded with the rocks of 
the Earth's crust. So it appears that as we use the coal formed many 
millions of years ago, potential coal is now forming for use many millions 
of years hence. 

Quaking Bogs. In the lakes and swamps of cool temperate and cold 
climates there flourishes a plant known as sphagnum moss. Sphagnum 
grows abundantly in the northern United States, Canada, and northern 
Europe, giving these northern swamps an aspect different from the 
swamps of the south. It readily grows outward on the water surface 
of a small lake, forming a floating mat which conceals clear water and 
black liquid muck beneath. When one walks on the mossy surface, 
the whole mass shakes and quivers; hence these bogs are called quaking 
bogs. Men and animals have not infrequently fallen through such 
unstable surface mats and have been lost in the quagmires below. 
Since the antiseptic nature of the bogs, deadly to bacteria, largely pre- 
vents the decomposition of organic matter, the bodies of animals en- 
tombed many thousands of years ago are dug up in remarkable states 
of preservation. In the state of New York alone the remains of more 
than 200 elephants which became mired during or after the Great Ice 
Age have been dug up from peat bogs. 

Economic Value of Swamp Lands. Peat is not the only valuable 
deposit of swamp origin. In lakes and swamps as well as in warm 
springs and even in the surface waters of the ocean live great swarms 
of microscopic plants called diatoms, which secrete minute shells of 
silica. When these organisms die, their tiny shells fall to the bottom 
and build up a white, porous deposit known as diat&maceous earth. Beds 
of this material, some of them hundreds of feet in thickness, are now 
excavated and used in the manufacture of dynamite, polishing powder, 
and other products. 


Certain valuable deposits of iron ore can be traced back to their 
in. lakes and swamps. Many areas of marsh and standing water 
today are inhabited by countless numbers, of microscopic organisms 
known as iron bacteria, living wherever the water contains dissolved 
iron. As an essential part of their life process, they secrete the iron 
from solution, convert it into insoluble form, and precipitate it, causing 
an accumulation of iron (usually in the form of limonite) on the lake or 
swamp floor. The bacteria, though minute, exist in such inconceivable 
numbers, and in successive generations work through such long periods 
of time, that large deposits of iron are gradually built up. These de- 

Fig. 123. Reclaimed land from the Dismal Swamp, Va. (IT. S. Geol. Surv.) 

posits where commercially important are known as one type of " bog - 
iron ores." They are particularly abundant in the glaciated regions 
of Europe, Asia, and North America, although they are now of very 
slight importance in the mining industry. 

The agricultural value of swamp land when drained is not to be 
minimized since swamp soils are highly fertile. The swamps and bog 
lands- in the United States have a combined area greater than the area 
of New England, and with proper draining much of this land could be 
reclaimed for farming. Drainage canals and ditches have been used 
for centuries in Flanders and Holland; and now in many parts of the 
United States large areas of wet land are being prepared in this way for 
agricultural use; Thus artificial means are being used to accomplish 
what in time would have been achieved by normal erosion more 
complete gradation of the land. 



1. Les Lacs; by Leon Collet. 320 pages. Paris, 1925. 
The most up-to-date work on lakes. 

2. The Lakes of Southeastern Wisconsin; by N. M. Fenneman. 1ST pages. 
Wise. Geol Survey, Bull. 8 (2nd edition), 1910. * 

Description of glacial lakes. 

3. Lake Bonneville; by G. K. Gilbert. 43S pages. U. S. Geol. Survey, Monogr. 
1, 1890. 

4. The Scientific Study of Scenery; by J. E. Marr. 361 pages. Chaps. 11 and 
12. London, 1920 (6th edition). 

A short popular discussion based largely on European lakes. 

5. Lakes of North America; by I. C. Russell. 125 pages. Boston, 1S95. 
A comprehensive popular account. 



Geologic Role of the Oceans and Seas. Marine waters cover 
about three-quarters of the globe. Their influence on the geologic 
history of the Earth's surface has been both direct and indirect. The 
importance of the oceans in regulating climate, for example, is well 
known. It is a remarkable fact that throughout the hundreds of 
millions of yeans of geologic history, the general temperature of the 
Earth has never fallen Mow the freezing point nor exceeded the boiling 
point of water. For this remarkable constancy we must thank the 
unflagging energy of the sun's warmth, but the oceans have also un- 
doubtedly served as a great stabilizer by storing up excess warmth 
against times of lessened solar radiation. Ocean currents carrying 
warm waters into the higher latitudes and returning the colder waters 
toward the equator also aid greatly in distributing the heat and softening 
the contrasts between climatic zones. Moreover, evaporation from the 
ocean's surface in the last analysis supplies all of the moisture borne by 
the winds to fall as rain and snow upon the lands. The intense aridity 
of the basins of the several continents, as the great desert of Central 
Asia, the deserts of Australia, the Sahara, and the Great Basin of the 
western United States, suggest what might be expected if the Earth 
had limited oceans. 

Through its waves and currents the marine water is an energetic 
agent of erosion, gnawing away relentlessly at the margins of all the 
lands. Moreover, the seas and the oceans are the final settling reservoir 
in which are deposited the sediments brought down by the rivers, as 
well as those produced by marine erosion. Upon the sea floor these 
sediments are shifted about to fill the depressions, or, in the shallow 
places, tossed up to build out the coast lines of the lands and ultimately 
to be compacted and cemented into sedimentary rocks. 

Relations of Oceans and Continents. If the waters were withdrawn 
from the ocean basins, we should see that the grandest relief features of 
Earth's rocky crust are not its mountain ranges but the uplifted con- 
tinental masses that lie as vast plateaus 3 miles above the enormous 
plains of the ocean floor. The naked face of the moon presents to us a 



spectacle of this sort, for its rnaria or " seas ?1 would be such in reality 
if there were water upon its surface. 

Presumably the continental masses stand high because they are made 
of lighter, granitic rocks and the oceanic areas are depressed because 
they are formed of heavier, basaltic rocks. On the moon, the depressed 
segments are relatively small; but Earth's oceans are far greater than 
its continental areas. It is apparently only a coincidence that there is 
just sufficient water on the Earth to fill the ocean basins brim full and 
place the shoreline upon the margin of the continents. 

Size of the Oceans. Not only do the ocean basins have three times 
the area of the continents, but they attain a maximum depth of more 
than 6 miles (35,410 feet) and an average of about 2-| miles (13,000 
feet), whereas the average elevation of the emergent continents is only 
about half a mile above sea level. So vast is the volume of oceanic 
waters that if the continents were planed down and leveled into the 
basins, a universal ocean would cover the entire Earth to a depth of 
more than If miles. It is important to realize, however, that even this 
depth is slight in comparison with the diameter of the Earth. If, for 
example, a globe 3 feet in diameter were dipped into water and then with- 
drawn, the film of wetness adhering to it would represent to true scale 
an ocean half a mile in depth, and if, in drying, the globe should warp 
by so small an amount as to lessen its diameter at any place by one- 
hundredth of an inch, the change would correspond to the depression 
of one of the major ocean basins. It is evident, therefore, that rela- 
tively trivial warping of the Earth's crust as a whole would suffice to 
deepen parts of the ocean basins and draw off the water from the con- 
tinents or, by elevating parts of the ocean floor, to cause an overflow 
of the lower land areas, with vast marine inundations. Whatever the 
cause, such changes have occurred many times in the geologic past. 
Some of them will be described in Part II of this book. 

At present the oceans more than fill their basins and flood the margins 
of all the continents. These slightly submerged borders of the conti- 
nental masses are known as continental shelves (Fig. 124). They are 
widened on the one hand through landward planation by the sea and 
on the other through the deposition of sediments swept toward the 
ocean by marine currents. They are broad along stable coasts where 
these processes have operated without interruption for long geologic 
ages, but commonly are narrow where young mountains have been 
formed near the continental margin. Thus, for example, the continen- 
tal shelf is 60 to 80 miles wide on the Atlantic coast off the Carolinas 
and only 10 to 25 miles, or even less, along the Pacific coast of California. 

Although it is customary to use sea level as a common datum of 


reference, it should be noted that the surface of the oceans is not a 
perfect sphere. Its polar diameter is about 27 miles less than the equa- 
torial and there are, in addition, local and irregular departures from the 
spherical surface due to the fact that the gravitational attraction of the 
continental masses draws the water up about the shores somewhat as 
surface tension draws up the margins of the water in a vessel. Thus 
the sea level is higher on the coasts than far out in the ocean and not 
as high on low coasts as on those bordered by high land masses, like 

Shelf Sea Sea level 

Fig. 124. Section through edge of continent into ocean basin. Vertical scale greatly 
exaggerated, causing slopes to appear much steeper than they actually are. 

the Andes. It is not improbable that such distortion of the sea level 
amounts, at its maximum, to several hundred feet. 

Oceans Versus Seas. In common usage the words sea and ocean are 
svnonymous, and we speak of deep-sea deposits, and of going to sea, 
when actually we have in mind the ocean. There is, however, a tech- 
nical distinction observed by scientists who apply the term ocean only 
to those five vast bodies of deep water that occupy the ocean basins 
proper and lie between the continental masses; namely, the Atlantic, 
the Pacific, the Indian, the Arctic, and the Antarctic. The seas, on 
the contrary, are relatively shaUo^^^ water that^lie 

upon the continental platforms as, for example, the North Sea or Hudson 
Bay. This technical restriction of the term sea accords with its original 
use by the peoples of northwestern Europe, who applied it to the North 
Sea and the Baltic in contradistinction to the Atlantic or outer ocean. 

The distinction is one of great significance because of the profound 
influence of depth on the processes that operate on the ocean floors. 
The seas are for the most part less than 600 feet deep, whereas the aver- 
age depth of the oceans is more than twenty times as great. The floor 
of the seas is a region of activity, touched by the warmth of sunlight 
and stirred by waves and currents that shift its sediments and bring 
oxygen and land-derived food to its teeming life. It has been the cradle 
of evolution for the lower forms of life and the scene of geologic processes 
of profound importance. The great oceanic floor, on the contrary, is a 
realm of dreary monotony. Except where submarine volcanoes or 
earthquake lines are active, nothing disturbs its quiet save the noiseless 
descent of wind-blown or meteoric dust or of derelicts of the surface 

S ANt> SEAS 171 

life miles abov6 ; and its denizens eke out an existence in utter darkness 
and cold. 

The Mediterranean " Sea " fits the definition of neither sea nor ocean, 
for it is of oceanic depth and yet is comparable in size to a sea and 
closely circumscribed by lands. The term mediterranean is therefore 
used in a generic sense to include such circumscribed bodies of deep 
water lying between continents. The Caribbean " Sea " is thus an- 
other mediterranean. 

Classes of Seas. The seas may be arranged, for convenience of 
reference, in three classes: marginal, epeiric, and relic. 

Marginal seas are those which lie upon the continental shelf and are 
more or less openly connected with the ocean. Where they are some- 
what delimited by projecting lands, various portions of the marginal 
seas bear distinct names; as the North Sea, the Yellow Sea, the Gulf 
of St. Lawrence, Cape Cod Bay, or the Gulf of Maine. The shallow 
water overlying the continental shelf of the central and southern At- 
lantic coast of the United States is just as much a sea, but it and others 
similar have not been named. 

The marginal seas fall naturally into two further subdivisions. Those 
lying upon the continental shelf proper are known as shelf seas. These, 
including the examples cited above, are shallow. The other group, 
known as funnel seas (Grabau), occupy strong structural depressions 
that trespass upon the continental border, like the Gulf of Lower Cali- 
fornia, the Bay of Bengal, and the Arabian Sea. They are far deeper 
than the typical shelf seas, especially at the outer ends where they are 
less sharply marked off from the ocean basin proper. They partake of 
the characters of mediterraneans but are less distinctly hemmed in by 

The epeiric seas [Gr. epeiros, a continent] are those that lie so far in 
upon the continent as to be largety land-locked. At the present time 
there are but two good examples, the Baltic Sea and Hudson Bay, but 
during certain past geologic ages, when the lower portions of the con- 
tinents became widely flooded, epeiric seas were of vast extent and great 
importance. In fact most of the sedimentary rocks upon the present 
lands were formed in epeiric seas. 

The relic seas are those like the Caspian, which, through crustal uplift, 
have become isolated from the oceans. They are, indeed, great brack- 
ish lakes and should be so called; but their former marine connections 
may be clearly indicated by the life they still harbor. For example, 
although the waters of the Caspian are now almost fresh, due to the 
inflow of great rivers, its fauna includes marine fishes, porpoises, and 
seals. It is known to have been connected with the Arctic Ocean as 



well a* the Mediterranean within rather recent geologic time. Lake 
Champlain i* also a relic sea, for it was flooded by marine waters from 
the Gulf of St. Lawrence a few thousand years ago, and fossil marine 
shells are abundant in places in the clay above its present shores. 

Depth Zones. The depth of the water is one of the most important 
factors in the marine environment for it conditions the amount of light 
and warmth so vital to marine life, and also the effectiveness of the 
waves and currents that determine the character of the sediments. 
From this point of view, the marine realm may be subdivided into 
several bathymetric or depth zones (Fig. 125). 

The littoral zone extends between extremes of high and low tides and 
includes, besides the actual shore slope, the mud flats, laid bare at low 

Littoral Zone 



Pelagic' Zone 

Fig. 125. To show the several depth zones of the marine realm. The vertical scale 
and the steepness of slopes are much exaggerated, and the full depth of the ocean is not 

tide, and the salt-water marshes, which are flooded only at highest tide. 
The entire width of the littoral zone is usually not over 2 miles and 
as a rule it is much less so that its total area is only about 60,000 square 
miles. In a few regions of exceptionally high tides or of rapid silting, 
however, the littoral zone reaches a considerable width. For example 
in the Bay of Fundy, where the average tide rises between SO and 40 
feet, the mud flats at ebb tide are 4 or 5 miles wide. In parts of the 
Mississippi River delta the salt-water marshes attain a width of fully 
25 miles. 

The environment of the littoral zone is more varied than that of any 
other part of the sea floor. Alternately covered by salt water and laid 
bare to the atmosphere, it may be baked in sunshine one hour, and 
deluged by the fresh water of a storm the next. Moreover, the waves 
break here with their greatest vigor, and wave and current action is 
strongest. At any one place the environment changes with the tides 


and storms from day to day, and it is equally variable from place to 
place along the shoreline, depending on the degree of exposure to the 
open sea. Accordingly the littoral region is the most difficult life 
zone, and relatively few kinds of animals live here though some, like 
the barnacles and crabs, exist in vast numbers. 

The neritic zone includes the shallow, sublittoral sea floor from the 
limit of low tide out to the margin of the continental shelf ; that is, to 
a depth of about 600 feet. It is known to the Germans as the Flachsee 
(flat sea) and, indeed, its floor is in general remarkably even, for it is 
mantled by the sediments derived from the land and shifted by the 
waves and currents into the deeper places, smoothing out the inequali- 
ties of its surface. The average width of the neritic zone being about 
75 miles, its seaward slope averages less than 10 feet to the mile, and 
it would appear to the eye as an utterly monotonous level plain. Only 
locally, and generally near the shore, does it reach an inclination as 
great as 50 feet to the mile. 

The total area of the neritic zone is about 10,000,000 square miles, or 
approximately one-fourth that of the land areas of the world. In past 
geologic ages, however, the zone has been greatly expanded, as the lower 
parts of the land areas were extensively flooded by epeiric seas. 

The neritic zone is a realm of change and activity. Waves and cur- 
rents keep the water in motion, salinity and muddiness vary from place 
to place, and the temperature changes with the seasons and varies with 
the latitude. Sunlight penetrates to the bottom. It is the most 
hospitable and stimulating environment for sea life and the teeming 
host of creatures that live here play an important geologic role. 

Beyond the edge of the continental shelf there is a somewhat steeper 
incline, known as the continental slope, that is fairly well defined from 
a depth of 600 to about 6000 feet, where it imperceptibly grades into 
the abyss of the deep ocean. The intermediate depths that cover the 
continental slope constitute the bathyal zone. Due to the greater depth, 
bottom currents are feeble here and no coarse sediment can be intro- 
duced; but fine, land-derived muds cover a vast expanse of about 
18,000,000 square miles, an area exceeding one-third that of all the lands. 

The abyssal zone includes the whole of the ocean bottom below a depth 
of 6000 feet. Its average depth is about 13,000 feet and in general it is 
monotonously flat, lacking the smaller relief features such as the hills 
and valleys of the land, Nevertheless on a large scale there are swells 
and depressions rising above and sinking below the general level of the 
ocean floor. The depressed areas, if more than 18,000 feet below sea 
level, are known as deeps and of these 57 have now been discovered. 
The deeps are of two rather sharply distinct classes: (1) the vast basin- 


like depressions with irregular borders that form the central portions 
of the several oceans; and (2) narrow and troughlike depressions, 
mostly situated near the continental margins and several of them paral- 
lel to marginal mountain ranges. A good example of the latter is the 
Tuscarora Deep, paralleling the Japanese Islands, with a depth of 
about 28,000 feet. These marginal deeps, or foredeeps, appear to be 
areas that have been depressed by breaking or sharp bending of the 
ocean floor to compensate for the uplifted marginal mountains. 

The greatest depth known is that of the Swire Deep, which lies along 
the eastern side of the Philippine Islands. About 50 miles east by 
north of Mindanao, a depth of 35,410 feet w^as discovered by the cruiser 
" Emden '' in 1927 and numerous other soundings in the vicinity show 
approximately 6 miles of depth. In the Atlantic, which in general is 
shallower than the Pacific, the Nares Deep, off Porto Rico, holds the 
known record with 27,972 feet. These great deeps of the ocean floor 
correspond, in area and in magnitude, to the highest elevations on 
the land, the oceanic depths attaining more than 6| miles below sea 
level and the loftiest mountain range, the Himalaya, about 5| miles 

The floor of the deep ocean presents an inhospitable environment 
for living things, since at this depth no sunlight ever penetrates the 
utter darkness; the pressure amounts to about 1 ton per square inch 
for each mile of depth; and the temperature is less than 4 C. Life 
is sparse and grotesquely specialized. In the absence of sunlight, no 
plant life exists and the animals are either carnivores or else scavengers, 
feeding on the dead organisms that settle down from the surface layers 
of the ocean. 

Over the abyssal floor, land-derived sediments are wanting, but 
everywhere there is a covering of peculiar deposits known as oozes. 
They are so soft and fine that water movements of one-half mile per day 
are sufficient to shift them on the bottom. These fine sediments and 
their sources are discussed on page 206. 

The surface waters of the open oceans constitute the pelagic zone 
or realm [Gr. pelagos, the open sea]. The organisms that live here 
enjoy the warmth and sunlight but must perpetually swim or float. 
As indicated above, they may contribute to the bottom deposits of the 
ocean basins. 

Composition of Marine Water. A barrel of sea water contains 
about 12 pounds of dissolved mineral salts. These dissolved salts con- 
stitute about 3J per cent of the weight of marine waters. Chemi- 
cal analyses from all parts of the oceans, and from different depths, 
show that the composition of the water in the open ocean is remarkably 


uniform, and that over 99 per cent of the dissolved mineral matter 
probably represents only a half dozen common salts, as follows: 

Sodium chloride, NaCl 77.8 per cent 

Magnesium chloride, MgCl 2 10.9 per cent 

Magnesium sulphate, MgS0 4 4.7 per cent 

Calcium sulphate, CaS0 4 3.6 per cent 

Potassium sulphate, K 2 SO 4 2.5 per cent 

Calcium carbonate, CaC0 3 0.3 per cent 

Minor constituents , 0.2 per cent 

100.0 per cent 

Among the minor constituents are traces of a surprising number of the 
chemical elements, including fluorine, boron, arsenic, nitrogen, phos- 
phorus, silicon, radium, copper, iron, lead, silver, and gold. Most 
of these occur in such small amounts, however, that they are of no direct 
geologic importance. 

Although common salt (NaCl) constitutes more than three-quarters 
of the dissolved mineral matter, there is little more than a trace of 
calcium carbonate (CaC0 3 ), the ratio of salt to calcium carbonate 
being almost 260 to 1. This relation is the more striking when it is 
recalled that river water, whence the oceanic salts have been derived, 
normally carries far more calcium carbonate than sodium chloride. 
The discrepancy is due, of course, to the fact that CaC0 3 is either used 
by organisms or chemically precipitated about as fast as it is delivered 
to the sea, whereas nearly all the salt (XaCl) that has been brought 
down to the sea since the beginning of time remains in solution. The 
total quantity of the dissolved salts is astonishing, for it amounts to 
about 32,000 million million tons, and if precipitated and crystallized 
into a bed of solid salt, would be sufficient to cover the whole of the 
United States to a depth of more than 1| miles. 

Silica and nitrogen, though present only in traces in marine waters, 
are of vast biologic consequence. The nitrogen is required as food and 
the silica as shell material by the single-celled plants known as diatoms, 
which float near the surface in incredible numbers, and form, to a large 
degree, the ultimate food supply for all the animals of the marine realm. 
Diatoms flourish and multiply rapidly if the supply of dissolved nitro- 
gen and silica is enriched but their expansion is checked by a falling off 
of these foods, so that the " pastures of the sea " may be said to depend 
on the supply of nitrogen and silica. The latter is one of the common 
minerals in river water and its rarity in the sea is probably the result 
of its use and precipitation by organisms. 


In addition to the mineral salts, marine water holds in solution oxygen, 
nitrogen, carbon dioxide (C0 2 ), and other gases. As a rough average, 
1000 cu. cm. of sea water holds in solution about 20 cu. cm. of air 
(oxygen and nitrogen) and 23 cu. cm. of carbon dioxide, both measured 
at standard temperature and pressure. The importance of these gases 
is out of all proportion to their quantity. All the animal life in the 
seas and oceans requires oxygen. Oxygen is necessary, moreover, for 
the oxidation and destruction of the carcasses of dead organisms on the 
ocean floor, and it thus aids in keeping the water clean and habitable. 
In places, such as the depths of the Black Sea, where defective vertical 
circulation fails to provide a sufficient supply of oxygen, the water 
becomes putrid and foul with the gases of decay, and only anaerobic 
bacteria are able to live. The entire ocean bottom would be equally 
forbidding if it were not for the bountiful supply of dissolved oxygen. 

Carbon dioxide is the primary food stuff from which green plants 
draw their sustenance and in turn build up the organic foods upon which 
animal life is dependent. Without it no life would be possible. At the 
same time it aids in the solution of the calcium salt, CaCOg, as explained 
in Chapter III. Since the solubility of C0 2 varies inversely as the 
temperature, the cold waters of the polar regions and of the deeper 
ocean bottom can hold more C0 2 than the warmer surface waters. It 
is this excess of C0 2 in the abyssal depths that causes the solution of all 
limy shells that settle there. 


Waves. The common movements of marine water include the 
waves, tides, and a variety of currents. Probably the most important 
and the most -direct of these in their geologic effect are the waves. 

The ordinary waves of the sea are generated by the wind blowing in 
irregular gusts and pressing unevenly upon the surface of the water, 
which Is thereby thrown into little undulations. Once formed, these 
undulations are maintained and increased by the pressure against their 
windward side and so are driven forward in endless succession. 

It is important to realize that the waveform travels ahead, independent 
of the movement of the water itself, just as waves may ripple across a 
field of standing grain though the individual stalks merely bow as each 
wave passes and then return to their original positions. Indeed, if the 
water actually rushed forward with the velocity of storm waves, the 
oceans would hardly be navigable. The path of movement of an indi- 
vidual particle of water is almost a closed circle (Fig. 126), for it rises and 
rides forward with the crest of the wave only to slide back into the next 


trough to its original position, a fact that can be observed by watching 
a bit of cork or driftwood as the waves pass under it. As a matter of 
fact,, due to the friction of the wind, it may not return quite to its original 
position but may be slightly advanced by each wave. In this manner, 
wave-formed currents are generated. 

The height (i.e., vertical elevation of crests above trough) and the 
length (i.e., horizontal distance from crest to crest) of the waves increase 
with the velocity of the wind, its duration in a given direction, and the 
length of " fetch " across open water. For this reason small bodies of 
water and protected embayments of the coast are never affected by 
great waves. A distance of 400 or 500 miles is sufficient to produce the 
greatest waves, however, since the wind does not blow steadily in any 
one direction over greater distances. 

Size and Velocity of Storm Waves. Observations have shown that 
storm waves in the North Atlantic commonly have a length of 400 feet 
and a height of 20 feet, but in times of exceptional storms they may 
attain a length of over 1000 feet and a height of more than 40 feet. 
Captain Tanner of the United States Navy has photographed a storm 
wave exceeding 50 feet in height. The velocity of the waves increases 
with their length and is about 25 knots, for example, when the waves are 
400 feet long and 40 knots for those 1000 feet long. 

When great waves run out of the storm they decrease in height and 
become more rounded, but they may continue for hundreds or thousands 
of miles as the long heavy undulations of the surface known as " ground 
swells." In this form the waves may attain a length exceeding 2000 feet. 

The wave motion decreases rapidly in depth as shown in Fig. 126 and, 
according to theory, the orbit of movement of a particle at a depth of 
one wave length should be only 1/534 of that at the surface. That is 
to say, in storm waves 20 feet high and 600 feet long, the surface particles 
of water will move through orbits having a diameter of 20 feet, and those 
at a depth of 600 feet through an orbit only f of an inch in diameter. . 
If, on the other hand, the wave be 450 feet long and 30 feet high, the 
orbit of movement of a particle at 50 feet down would be 15 feet, at 100 
feet down 7.5 feet, at 150 feet down 3.75 feet, and at 450 feet down about 
half an inch. Even great storm waves, therefore, cannot disturb the 
bottom below a depth of a few hundred feet. Strange as it may seem, 
the depth to which the wave is effective depends rather on its length 
than its height. Observations on the depth to which the sea floor is 
affected by waves are rather conflicting, undoubtedly because the depth 
affected varies from pla^e to place, depending on several factors. Ob- 
servations by divers off the south coast of England have shown that 
pebbles are rolled about, at a depth of 50 feet during a storm, and Fol 


records that, at a depth of 100 feet, he was tossed back and forth by the 
oscillation of the water on the bottom when the ground swells were 
running at the surface. Cobbles weighing a pound or more are some- 
times washed into the lobster pots at a depth of 180 feet off Land's 
End, England, and wave ripples have been recorded in very fine sand 
at a depth of 617 feet near Madagascar. The depth at which fine, soft 
mud reposes on the sea floor may be considered as a good test of 
the depth of wave disturbance, since mud could hardly remain per- 

Fig. 126. To show the decrease of wave motion with depth. The circles show the 
relative size of the orbits of motion at their respective depths as the wave passes. Ac- 
cording to theory the diameter of the orbit of movement of a particle is reduced by one- 
half for each successive increase of one-ninth a wave length in depth. (After Johnson.) 

manently on a wave-disturbed bottom. Judged by this criterion, the 
sea floors off exposed coasts are generally affected by waves to a depth 
of 200 or 300 feet, and exceptional waves move fine sediments to the 
edge of the continental shelf, a depth of about 600 feet. 

When a wave passes into shallow water, an important change takes 
place in its form and behavior. The wave becomes higher and shorter 
and its front side steeper and more deeply concave until the crest arches 
forward, unsupported, and collapses with a roar. These collapsing 
waves or breakers form the surf, so common a feature along coasts. 
The height of the wave determines the? depth at which it will brejak, and 
thus small waves break in shallow water, near shore, whereas the great 
storm waves break where the water is from 10 to 20 feet deep. 

Tides. Due to the attraction of the moon, modified by that of the 
sun, the ocean surface is lifted into two vast but low tidal bulges, one 
on each side of the Earth. These bulges remain fixed with respect to a 



line extending to the moon, but since the Earth in its daily rotation 
turns under them from west to east they seem to run round the Earth 
from east to west. 

In the open ocean, the surface merely rises a little and then subsides 
again as each tidal bulge passes; but where the bulge impinges against 
a coast the water is dragged forward, piling up on the shore and then 
receding, thus producing the familiar phenomenon of the tides. The 
height of the tide is determined largely by the configuration of the coast. 

^^r^^^^>^^^^ ;-^*w *,-. 

Fig. 127. High and low tide in the Bay of Fundy. Note the same bridge and sailing 
vessel in both views. Port Williams, Nova Scotia. 

On open, exposed coasts it is not more than 6 or 8 feet, and in protected 
embayments as, for example, the Gulf of Mexico, it is only 1 or 2 feet; 
but in estuaries that open out toward the advancing tide, the water 
rises higher as it converges, bringing about exceptional conditions like 
those in the Bay of Fundy, where the tide rises normally 30 to 40 feet 
and exceptionally as much as 50 feet. (Fig. 127.) 



The immense bodies of water thus moving in and out of bays and 
estuaries and along the coast every 6 hours, produce strong tidal cur- 
rents, which, like rivers, have a two-fold geologic function in that they 
both erode and transport. The work done by the tides in scouring the 
bottom and transporting material along some coasts is very great. 
The rise and fall of the tides indirectly aids in the attack of the waves 
on the land by increasing the vertical range of their contact with the 

Currents. There are several types of marine currents, only the 
chief of which need be considered here. Surface currents are produced 
by the friction of the wind when it blows persistently in one direction 
for a time. Where they strike the coast and, with the aid of the break- 
ing waves, tend to pile the water up on the shore, there is a compensating 
bottom current, the undertow, flowing seaward (Fig. 128). On a straight 

]fig. 128. To show the relation of the wave-formed surface current (5) and the under- 
tow (U) when the waves are coming directly onshore. 

coast, with a gently shelving sea floor,, the undertow is so diffused as 
to be very gentle, but where irregularities in the bottom constrict the 
returning current to more definite channels, the undertow may be a 
menace to swimmers and a powerful agent in sweeping sediments sea- 
ward into deeper water. Where the waves strike the shore obliquely, 
the run of the water due to successive impulses of the waves generates 
a current parallel to the shore, known as the littoral or shore current 
(Fig. 129). Such currents sweeping material along the coast are respon- 
sible for the formation of beaches, spits, and bars. The fine sand of 
Daytona Beach, Florida, for example, comes not from the interior of 
that limestone country, but along the shore from far to the north, where 
the rivers of Georgia and the Carolinas have brought it down to the 
sea out of the southern Appalachians. 

Tidal currents are strong where the tide sweeps in and out of deep 
estuaries or bays. Where the shoreline is oblique to the east-west path 



of the tidal bulge, there may be a strong component of the tidal current 
running parallel to the shore. This is true, for example, along the west- 
ern coast of Newfoundland, where the ebbing tide sets to the northeast 
through the Strait of Belle Isle with a velocity of 4 or 5 miles per hour, 
a rate that compares well with the flow of large rivers. 

The more general ocean currents have for the most part only an in- 
direct geologic effect by modifying the climate of the lands, for they 

Fig. 129. To show the relation of the wave-formed surface current () to the littoral 
current (L) and the undertow ( C7) when the waves are oblique to the shore. 

rarely touch the shores or the bottom with sufficient velocity to erode 
or transport sediments. On the other hand, they have a distinct geo- 
graphic value in drifting organic waifs to strange lands. These cur- 
rents are generated by several different factors. The unequal heating 
of the ocean by the sun in tropical and polar regions would establish a 
slow general circulation of its waters through convective movements. 
This action, however, is controlled, hastened, and magnified by the 
wind belts of the Earth, described in Chapter III, and by the disposition 
of land and sea. Driven by the trade winds, there is, in the equatorial 
regions of the Atlantic as well as the Pacific, a broad current moving 
westward along the surface. When this strikes the continental coasts 
it divides, one part turning northward, the other southward; and each, 
circling, returns to the equatorial belt, thus making in each ocean a 
vast eddy, one north, the other south of the equator (Fig. 130). In the 
same manner there is a circling movement in the Indian Ocean. In 
the center of each ocean are more stagnant areas, that in the North At- 
lantic being known as the Sargasso Sea. When these broad slow move- 
ments, which are known as drifts, approach the coasts, the water tends 
to accumulate, and where confined by the configuration of the land is 
hastened in its motion, giving rise to streams. Thus in the North 



Atlantic, the equatorial current striking the north coast of South Amer- 
ica is deflected northward. A part enters the Caribbean Sea and 
the Gulf of Mexico, whence it issues through .the greatly confined 
Straits of Florida and passes northeast into the Atlantic as the well- 
known Gulf Stream. In the straits this current averages 72 miles a 
day and during the summer and winter months sometimes rises to 
120 miles a day; but as it spreads and approaches mid-ocean the ve- 
locity diminishes greatly, falling finally to 10 miles a day. 

As this drift approaches the shores of Europe, it divides and one part 
turns southward to pass along the coast of Africa, and so to rejoin the 

Fig. 130. Map showing main ocean current and drifts. 

westward equatorial drift. Another portion passes northward into 
the Arctic Ocean, and to balance this a cold current comes down from 
the western coast of Greenland, past Nova Scotia and New England, 
and gradually passes under the warm surface of the Gulf Stream to 
sink into the abyss. 

In a similar way, in the North Pacific a current moves along the coast 
of Asia, then eastward and finally southward along the western coast 
of North America. It is known as the Japan current. These warm 
currents, moving into northern latitudes, have a great effect upon cli- 
matic conditions in the lands whose shores they strike. Thus the ocean 
currents, like the atmosphere, by taking part in the general circulation 


on the surface of the globe are great distributors of heat. They carry 
warmth into the polar seas and, returning as oold currents, bring with 
them reduced temperatures and icebergs to be melted in warmer regions. 
Were it not for their agency, ice would continually increase in the 
polar areas. The indirect effect of the ocean currents on geological 
processes is therefore important, although these currents perform little 
direct geological work. 

Differences in temperature may generate ocean currents, the denser, 
colder water tending to move toward the lowest places and to displace 
the warmer water toward the surface. For this reason there is a slow 
and general movement of the polar waters along the bottom into all the 
deeper ocean basins, which consequently have a temperature approxi- 
mately at the freezing point of surface fresh water. In the depths of 
the Pacific, Indian, and South Atlantic, for example, the temperature 
ranges between 1 C. and 4 C. If there were no cold polar regions, the 
ocean bottoms would not be so cold. 

Differences in salinity may also set up vertical currents, because the 
density of the water increases with its salinity. The lighter fresh water 
of many great rivers extends far out over the surface of the salt water, 
mingling gradually by diffusion. On the other hand, in regions of 
excessive surface evaporation the upper water of the ocean may become 
concentrated enough to form downward currents compensated by the 
rise of cold water from the depths. Professor T. C. Chamberlin de- 
veloped the ingenious hypothesis that, at times in the geologic past, 
this factor may have reversed the present' vertical circulation in the 
oceans. At present the influence of the temperature dominates and the 
general vertical circulation carries the cold waters down and equator- 
ward, the warm waters moving poleward at the surface until their warmth 
is gradually lost by radiation. Chamberlin suggests, however, that 
at times of warmer polar regions the effect of salinity may have dom- 
inated, the evaporation in equatorial regions then causing the warm 
surface water to settle and be crowded poleward along the bottom, 
whence it emerged with its stored heat in high latitudes. In such an 
event, the polar regions would be much more effectively warmed and the 
tropical heat ameliorated by the surface flow of cooler water, the polar 
regions remaining shrouded in fog due to the ascending warm water. 
While this speculation offers a possible explanation of the more equable 
world climates of many past geologic ages, it must be confessed that 
there is no direct evidence that can be cited in its support. 

Ascending currents bringing cold water from the depths near some of 
the tropical shores render the climate of the coastal belt surprisingly 
cool. This is true of the coast of Peru, and likewise of parts of the north- 


west coast of Africa as, for example, the coast of Morocco which lies 
almost at the edge of the tropical desert but is refreshingly cool. It is 
partly because of the ascending currents along the Pacific coast of 
California that San Francisco and Los Angeles enjoy summer tempera- 
tures so far below that of Washington, D. C. ; and the Virginia Capes. 


Erosive Processes. The surface of the seas, like an ever-moving 
horizontal saw, is ceaselessly cutting and gnawing away at the lands. 
This it does in several ways, producing a variety of features that are 
worthy of consideration. 

Sea water has more or less solvent effect on rocks, tending to disinte- 
grate them, and thus aid in their destruction; but the chief factor in 
marine erosion is the mechanical attack of the waves. The force of 
impact of large waves due to the sheer weight of tons upon tons of 
surging water may be sufficient to erode unconsolidated sediments 
rapidly or to move enormous blocks of stone. According to theory, the 
pressure exerted by a wave 10 feet high and 100 feet long is 1675 pounds 
per square foot, and of a wave 12 feet high and 200 feet long 2436 
pounds per square foot; and " great ocean waves, if we assume a height 
of 42 feet and a length of 500 feet, should produce a pressure of 6340 
pounds per square foot " (Johnson). These theoretical calculations 
accord well with actual dynamometer measurements, which on the 
coast of Scotland showed the average force of the summer waves to be 
611 "pounds per square foot and the average for the winter months to 
be 2086 pounds with extremes up * o 6083 pounds. The damage done 
by storms to harbors and breakwaters bears further testimony to the 
force of the waves. During a great storm at Wick, Scotland, in 1872, 
a solid mass of stone and concrete weighing 1350 tons was torn from its 
place at the end of a breakwater and dropped unbroken inside the pier. 
When great waves strike against a sea wall or solid cliff, they not un- 
commonly dash up to heights of 100 feet or more. At the lighthouse on 
Tillamook Rock, on the exposed coast of Oregon, the water of the waves 
was thrown over 200 feet above the sea during a storm in 1902, and 
during the winters of both 1912 and 1913 the impact of the waves broke 
panes of plate glass in the lantern of the same lighthouse at a height of 
132 feet above the sea. It is not surprising, therefore, that the faces 
of bold cliffs are shattered, and softer materials rapidly dislodged, 
simply by the impact of the waves. (Fig. 131.) 

Hydraulic pressure may be a powerful agent of the waves as well, 
for most rock masses have crevices or larger cracks in them, and the 



air or water in these, driven violently in by the impact of the waves, 
acts as a wedge, disrupting them and dislodging large pieces. In this 
way heavy masonry is often torn asunder. Moreover the water, rush- 
ing into cavities and suddenly retreating, leaves a partial vacuum, 
which tends to suck away portions of the roof and sides, and the con- 
stant repetition of this action aids in forming sea caves, blowing holes, 
and spouting rocks, so frequently seen on rocky coasts. 

The chief eroding action, however, is accomplished by abrasion, and 
in performing this work the waves use as tools the dislodged material, 

Fig. 131. Storm waves breaking against the sea wall at Hastings, England. (After 
Johnson. Photo by Judges.) 

and also that which falls from above and tends to form a talus at the foot 
of the sea cliff. The constant striking and grinding, not only of sand 
and gravel, but even of heavy boulders, render the waves formidable 
agents of destruction, through whose work even the hardest rocks are 
worn away. The ineffectiveness of clear water, aside from the me- 
chanically disrupting processes, is strikingly shown in places on the 
coast of Norway, where headlands extend into water too deep to be 
affected by coastal debris. The rock surfaces, smoothed and furrowed 
by former action of glacial ice, still retain these characteristic features, 
though subjected to the. constant washing of the waves. 

In the process of abrasion the material used by the waves is itself 
ground up and reduced to sand and silt. It loses its angular character 
and takes on the rounded form characteristic of coastal debris sub- 
mitted to chafing by the waves. 


Transportation of Sediment. If wave erosion is to continue, the 
ground-up rock debris must be removed, like sawdust from the track 
of the saw, in order that fresh rock surfaces may be exposed to attack; 
otherwise the fine material would act as a buffer to receive the waves 
and prevent further erosion. This removal is done by the undertow, 
which takes the debris out to sea, and also by tidal and littoral currents, 
which carry it away. Were it not for the aid of these agencies, wave 
erosion would cease except where the sea encroaches on a sinking land 

As in the case of streams, the products of erosion may be transported 
in the sea, either in solution, in suspension, or by bottom rolling. But 
unlike stream transportation, the currents in the sea are aided greatly 
by the waves. 

Due to the orbital motion of the water in an ordinary wave of oscil- 
lation, a grain of sediment on the bottom tends to be lifted and carried 
slightly forward as each wave crest passes, but is carried down and back 
to its original position by the trough of the wave, so that no actual trans- 
portation is accomplished by the wave itself. But when the particle 
of sediment is lifted free of the bottom, even the gentlest current can 
deflect its fall; in this way gentle currents, aided by the repeated lift 
of the waves, can sweep along sediments that they alone would be power- 
less to move. 

When waves reach shallow water and begin to drag heavily on the 
bottom, their movement ceases to be one strictly of oscillation, for the 
water then tends to roll ahead as a wave of translation which, like a 
strong current, drags the bottom material forward with it. Since the 
waves drag bottom only when approaching the shore, they tend to 
transport the bottom sediment landward. This tendency is opposed, 
however, by the undertow and by gravity, both tending to shift the 
sediment down the seaward slope. These seaward forces act continu- 
ously and are able, therefore, to overcome the stronger but intermit- 
tent surges of the waves. Although the sediment is dragged back and 
forth with each passing wave, the net movement of the fine material 
is seaward. On the other hand, the undertow may be unable to trans- 
port coarse material which the breaking waves drag forward, and when 
this is true the coarse sediment migrates landward even though the 
fine material is being shifted seaward. Murray has recorded, for ex- 
ample, that stone ballast discharged from ships near the British coast 
in water as much as 60 feet deep, has been thrown up on the beach 
during a subsequent storm. This factor keeps the coarse sediment 
concentrated at the shore, where it is known as the beach shingle. 

Even fine sand is thrown shoreward to form a sandy beach where 


loose sediment is sufficiently abundant, the waves carrying forward 
more of the material than the undertow can remove. The beach is but 
a temporary accumulation of sediment in transit, however, for there is 
a constant loss to the undertow and shore currents that spread the 
material seaward. 


In their attack on the shore the waves and currents carve erosive 
features in some places and in others build up deposits of the loose 
sediment. The nature of the shore features thus produced at a 
given place depends partly upon the character of the land surface 
against which the marine agents have to operate and partly upon the 
length of time they have been at work. As a land surface under the 
influence of stream erosion passes through a cycle of physiographic 
changes from youth to maturity and old age, so also the shoreline 
passes through a predictable series of changes as the marine cycle pro- 
gresses, and many of the shore features, like those of a degrading land 
surface, are temporary features in the physiographic cycle. 

The marine cycle is inaugurated ordinarily in one of two ways; either 
(1) by an uplift of the sea floor against which the waves start their 
attack de now or (2) by a submergence of the coast which brings the 
sea into contact with a former land surface. The initial form of the 
shore on an emergent sea floor is utterly different from that of submerged 
land and the sequential forms developed from the one are so distinct 
from those of the other that each type of shore must be described 

The Shoreline of Emergence. When a sea floor emerges by gentle 
uplift it forms a nearly flat coastal plain mantled by unconsolidated 
sediments. The plan of this shoreline is at first very simple, since it 
marks the intersection of this flat surface by sea level. The water is so 
shallow for a considerable distance from shore that the waves drag 
heavily upon the bottom, picking up the loose sand and gravel, which 
they carry forward to the line of breakers. Here the loose material 
is dropped again as the wave spends itself, gradually building up a 
narrow submarine ridge parallel to the shore and just within the line 
of breakers. Storm waves eventually build the deposit above sea level, 
forming a long low sandy island parallel to shore (Fig. 132, A). This 
structure is known as an offshore bar or barrier beach. The narrow body 
of shallow water lying between the offshore bar and the shore is a lagoon. 
Large lagoons are also called sounds but not all sounds are lagoons. 
Ordinarily the sweep of the tide in and out of the lagoon keeps inlets 


througk the barrier open to the sea. The Atlantic and Gulf coasts 
of the United States display these features in grand development 
(Fig. 143). The width of the lagoon depends on the seaward slope of the 
bottom and the size of the storm waves. It is commonly a mile or more 
and may be several miles. At Cape Hatteras the offshore bar is about 
20 miles from the mainland but along the east coast of Florida it is near 
shore and the lagoon is known as Indian River. The city of Galveston 
is built on an offshore bar and its tragic flood of 1900 occurred when 
high seas, driven by a hurricane, rose 15 feet above their normal level. 

The sea floor outside the barrier is gradually excavated by the waves, 
supplying the material of which the barrier is made. As its depth in- 

Fig. 132. Showing youthful stages in the development of a shoreline of emergence. 
The offshore bar is stippled in section and the lagoon filling is vertically lined. A, early 
youth, showing the formation of the offshore bar; J3, a later stage showing partial filling 
of the lagoon ; C, a still later stage in which the offshore bar has been prograded toward the 
shore and the lagoon has become a fresh- water marsh; D, late youth; E, early maturity, 
the offshore bar and lagoonal deposits having disappeared. Vertical scale exaggerated. 
(After Davis.) 

creases the waves are less impeded by friction on the bottom and are 
finally able to break with force against the barrier itself, eroding ma- 
terial from its seaward side. A part of this debris is carried out to 
sea, and during storms a part of it is thrown over the bar into the edge 
of the lagoon. In this way the barrier is gradually prograded shore- 
ward. Meanwhile the lagoon, rarely more than 20 feet deep at the 
start, tends to be filled by sediment washed into it from the land, as 
well as by that thrown over the barrier. Where the water is shallow 
enough, vegetation thrives and adds its quota to the accumulating 
sediment (Fig. 132, B). As the barrier is gradually driven shoreward the 
lagoon becomes a salt-water marsh with open tidal channels (Fig. 132, C 
and D). Eventually both barrier and lagoonal deposits are completely 
cut away as the sea bottom is excavated up to the shoreline (Fig. 132, E) 
and then for the first time the waves begin their attack upon the land. 
During these early stages of the cycle, the bottom profile near shore 



has been gradually changing from the simple slope of the original sea 
floor to the compound curve represented in Fig. 133. There is a concave 
portion near shore where the bottom is being eroded. It extends sea- 
ward as a nearly flat wave-cut bench or terrace to the line of depth where 
wave cutting is ineffective. The depth of the wave-cut terrace depends 
on the size of the prevailing storm waves and the supply of sediment 
they have to move, but it is generally between 10 and 20 feet near shore 
and gradually deepens seaward through abrasion by the sediment that 
is shifted across it. Beyond the zone of cutting the sediment is dropped 
and gradually built out as a submarine embankment or wave-built 
terrace, which continues the nearly flat slope of the wave-cut terrace. 

Fig. 133. The wave-formed shore profile of equilibrium, concave near the shore where 
the sea is cutting and convex farther out where sediments are accumulating. The beach 
is represented by solid black and the sediments of the wave-built terrace are stippled. 
Vertical scale exaggerated. 

The bottom is convex over the outer part of this deposit and concave 
further out where the sea floor drops away to the normal bottom slope. 
The curve shown in Fig. 133 is the profile of equilibrium and its general 
character will be maintained to the end of the marine cycle, though the 
curve will gradually flatten out as the sea cuts farther inland and the 
waves waste more and more of their energy in bottom friction while 
crossing the wide neritic zone. 

Where the waves are cutting into solid rock the wave-cut terrace is 
rock-floored, though it may be more or less mantled with debris in 
transit from the shore; but where the attack is upon unconsolidated 
sediments the wave-cut terrace obviously has a floor of loose material. 

As the waves concentrate their attack at sea level, cutting horizon- 
tally into the sloping land surface, they tend to undercut the land. 
The overhanging material is gradually dislodged and falls down and 
thus the coast comes to be terminated in a sea cliff as in Fig. 132, E. 
The nature of the sea cliff depends in part on the material attacked by 
the waves. In firm rocks it may be very steep or even overhanging, 
but in loose material like sand, slumping and sliding keeps pace with 


undermining and the sea cliff is a slope equal to the angle of repose of the 
loose material. The form of the sea cliff is also influenced by the relative 
rate at which the shore material yields to weathering and to wave cut- 
ting. For example, clay is easily cut by the waves and commonly 
presents bold bluffs in spite of its softness, whereas hard granite may 
succumb more rapidly to weathering than to the attack of the waves, 
especially where it is much jointed and exposed to frostwork. The 
hard rock may thus present a sea front of low slope whereas soft rock 
may form cliffs. 

The debris that falls from the sea cliff or that is eroded at sea level 
(as well as that which is introduced by streams) is shifted back and forth 

by the waves and currents and the 
water level t finer material is carried seaward with 
the undertow. The coarser sedi- 
ment is kept at the shoreline, how- 
ever, to form the beach, a low ridge 
Fig. 134. Section of a beach. (After o sediment thrown up bv the waves 

Gilbert.) , , . A . J ,. , . 

along the shore. As indicated in 

Fig. 134, the beach is only a thin veneer on the bedrock floor, extending 
landward to the limit reached by the greatest storm waves. In cross 
section it is convex upward at the shore, where it is thickest. It ex- 
tends for a short distance under the water, thinning gradually and 
merging into the sheet of sediment that partly mantles the rock-cut 
bench while in transit out to sea. The upper margin of the beach is 
generally marked by a belt of coarser material thrown up by the heavi- 
est storm waves and this grades into finer material toward its seaward 
margin. In places where the shore zone is formed wholly of fine sedi- 
ment the beach is made of sand alone. 

If the rock forming the sea cliff is cut by vertical joints, the waves 
tend to widen the joints and quarry out some of the blocks, leaving 
others standing isolated as stacks (Fig. 135). If the joints are more 
irregular and intersect before reaching the summit of the cliff, sea caves 
are formed at the base of the cliff. As the waves break into the mouth 
of a sea cave they exert a heavy hydraulic pressure on its sides and roof, 
followed by rarefaction of the air as the wave recedes. By this action 
the roof of the cave may be excavated through to the surface so that 
spray will dash up through the roof with each breaking wave. Such 
caves are known as blow holes or spouting caves. Narrow promontories 
or islands are commonly undercut by the development of sea caves to 
form sea arches. 

Obviously the greatest inroads are made by the sea where the rocks 
are weak and the slightest progress where they are resistant. For this 



Fig. 135. Sea cliff and stack, the latter a remnant of former land now eroded away. 
Coast of Wales. (Geol. Surv. of England and Wales.) 

Fig. 136, - A bay with a curving beach. Conception Bay, Newfoundland. (Walcott.) 



reason a simple shoreline formed of heterogeneous rocks gradually de- 
velops irregularities as erosion progresses; the resistant masses stand 
out to form headlands or promontories and the weak rocks recede to 
form coves or bays (Fig. 136). If the rocks are composed of parallel 
layers or beds and their edges are exposed to the waves, the weaker, 
softer layers are rapidly worn away and the harder beds, left unsupported, 
break away in blocks. If the beds are horizontal the harder layers may 
project for a time as table rocks with cavities under them, as on the coast 
of Lake Superior. If the beds are vertical, or inclined, but with edges 
exposed, the hard layers stand out like columns or ribs (Fig. 137). If 

Fig. 137. Irregular coast and sea cliff produced by erosion in nearly vertical rock 
strata. Pembroke, Wales. (Geol. Surv. of England and Wales.) 

the face of the beds is towards the sea, erosion is slower because the 
hard layers form an apron, or wall, to protect the soft layers behind 
them. If the rock masses are homogeneous and hard, like trap or 
granite, the irregularities are largely determined by joints and by the 
arrangement of these with respect to the sea front. Thus a bold coast 
facing the sea is likely to show many minor irregularities of topography. 
As one notes the delicate adjustment of the shore configuration to the 
resistance of the rocks he might be inclined to think that all the irregu- 
larities of the shoreline are due to erosion. Such, however, is far from 
true for there is an effective limit to the depth inland to which embay- 


ments may be cut by the waves. For example, the deeper the bay the 
more completely the defending headlands protect it from all oblique 
waves. Experience shows that large storm waves do not enter far into 
deep embayments. Moreover the debris torn from the headlands tends 
to be swept into the protected bay-heads to form beaches that mantle 
and protect the shore. The great and deep embayments of the coast, 
therefore, result not from marine erosion but from crustal warping 
or the drowning of river valleys. 

In the physiographic development of the shoreline of emergence, 
the stages illustrated in Fig. 132 are all preliminary to the real attack 
against the land. They represent the youthful stage of the marine cycle. 
Following the disappearance of the offshore bar, the sea launches its 
drive with full vigor against the shore. This is the beginning of the 
stage of maturity. The sea cliff is now formed and the simple shoreline 
rapidly develops irregularities, until the plan of the shore is adjusted 
to the forces of attack and to the resistance of the rocks. Erosion 
then proceeds gradually along the whole line. The final stage of old 
age is attained only when the sea has cut so far inland that the waves 
spend most of their force in friction over the shallow bottom and make 
but a feeble attack upon the shore. After this stage is reached, further 
retreat of the shore due to marine erosion proceeds very slowly. 

The Shoreline of Submergence. The initial form of a shoreline of 
submergence is marked by extreme irregularity, since the marine water 
comes to rest against an eroded land surface, entering far up the valleys 
and forming deep irregular embayments between sloping headlands. 
Glaciated valleys thus drowned become fiords and the lower courses 
of the normal stream valleys become estuaries or deep bays. Hard-rock 
islands are likely to be abundant near the shoreline, representing isolated 
hills of the drowned land surface. 

The youthful stage of the cycle of erosion of such a shoreline is marked 
by lack of adjustment of the erosive forces to the plan of the shoreline. 
The waves must concentrate their attack chiefly on the offshore islands 
and the headlands that defend the deeper embayments. These are 
quickly cliffed, while the bay-heads tend to be silted up by the streams 
that enter them. In their fresh attack against the land surface, the 
waves quickly indent the weaker spots, developing sea caves, archeSj 
and stacks and forming little bays even in the headlands. The irregu- 
larities thus developed are on a small scale and are only incidents in the 
general process of straightening the shoreline, but they give to it a 
crengulate plan characteristic of early youth. 

JThree youthful stages and one of early maturity of the submergent 
shoreline are indicated in Fig. 138, It will be noted that they tend toward 





a simplification of the extreme irregularities of the initial shore. The 
first stage is marked by cliffed headlands. Since the surface is drowned, 
deep water comes close to the headlands and the debris first eroded 
sinks beyond reach of the waves so that no beach is produced. Grad- 
ually the headlands are cut back and the debris supplied by their erosion 
accumulates until a deposit is built up into the zone of wave action. 
A beach is then formed at the foot of the cliff. Littoral currents do not 
enter deeply into the estuaries and as they sweep past the truncated 
ends of the exposed headlands they carry part of the beach material 
laterally into deeper water where it settles, building out a submarine 
embankment. By continued action of this process the embankment is 

Fig. 139. A curved spit, or hook. Duck Point, Grand Traverse Bay, Lake Michigan. 
The left end, beyond the edge of the view, is attached to a point of land. (U. S. Geol. 

built out across the adjacent embayment very much as an artificial 
fill is made across a valley by the dumping of dirt at the end of the fill. 
Incoming waves will throw some of this material upon the embankment 
and build it above sea level. It then appears as a low sandy or gravelly 
promontory built out from the beach. If nearly straight this structure 
is a spit. Another current crossing the end of a spit may sweep the 
material aside, developing a curved or hooked spit which is known as a 
hook (Fig. 139) . If the deposit is extended until it closes, or nearly closes, 
the mouth of an embayment it is known as a bar (Fig. 140). Spits and 
hooks tend to form wherever the littoral currents sweep sediment away 
from the beach, whether it be at the end of a promontory or merely of 
a blunt cape. Bars tend also to form in the sheltered water between 



islands and the shore, resulting in land-tied islands known as tombolos 
(Fig. 141). Sea caves, arches, and stacks are especially characteristic of 

these early youthful 
stages of the submer- 
gent shoreline. 

While spits, hooks, 
and bars are forming 
about the retreating 
headlands the inner 
ends of the embay- 
ments tend to be filled 
by deltas where 
streams enter these 
protected bodies of 
deep water. 

Eventually, as in 
Fig. 138, C, the shore- 
line is much simplified 
as the land-tied islands 
are cut away and the 
headlands truncated 
and united by more or 
less continuous bars. 
The bay-heads become 
largely filled with sed- 
iment and pass into 
Calif: (u. s. sa it_ wa ter marshes 
like those that are so 
common along the New England coast at the present time. 

When finally the headlands have been completely cut away and the 
bay-head fillings are all removed, as in Fig. 138, D, the youthful stage 
of submergent coast is completed and maturity has begun. The waves 
are now at work along the entire coast, the plan of the shoreline is ad- 
justed to the resistance of its rock masses, and the bottom has developed 
the profile of equilibrium, concave over the wave-cut bench and convex 
over the wave-built terrace. The deeper irregularities in the sea floor 
have been filled with sediment. From this stage on the development is 
identical with that of the shoreline of emergence. 

Shorelines Produced by Interrupted Marine Cycles. Completion 
of the marine cycle would require stability of the land and a constant 
sea level for a very long time. A relative uplift or depression of the coast 
will interrupt the cycle and superpose a new cycle on the old. 

Fig. 140. A 

bar closing Mono Bay, 
Geol. Surv.) 



Fig. 141. A tombolo or land-tied island. Bay of Fundy, Nova Scotia. The left 
side of the connecting bar is not fully shown. 

Fig. 142. An elevated wave-cut terrace north of Port Harford, Calif. Old stacks 
rise above the terrace. (U. S. Geol. Surv.) 



If uplift of the coast is intermittent, the wave-cut bench formed during 
a period of stationary sea level may be lifted above the sea to form a 
marine terrace paralleling the shore (Fig. 142). Such terraces are strik- 
ingly developed for long distances along the southwest coast of New- 
foundland. At Port au Port, one of them bevels evenly across the edges 
of steeply tilted strata and attains a width of more than 2 miles. Al- 
though originally carved by the waves, it is now 55 feet above sea level. 
Abundant marine shells occur in the clay that mantles another terrace 
lying just 100 feet above sea level along the western end of Cape St. 
George, Newfoundland. Farther north in both Newfoundland and 
Labrador, there are remnants of older and more eroded marine terraces 
at several elevations up to about 500 feet above the present sea. Along 

the coast of California, marine 
terraces have been recognized up 
to 1600 feet above sea level. 

The landward margins of some 
marine terraces are marked by an- 
cient sea cliffs accompanied by 
other shore features such as stacks 
and sea caves. Such uplifted sea 
caves are used as shelters by in- 
habitants of the bold coast of Fife- 
shire in Scotland (Fig. 197) . Char- 
acteristic beach deposits also mark 
the position of some old shorelines, 
and where uplift has been intermit- 
tent, a series of beaches may lie 
one above another as they do along 
the coast of southeastern Labrador, 
where the land has only recently 
recovered from its depression by 
Fig. 143. Map of the coast of North the weight of the Pleistocene ice 

Carolina showing an association of drowned 
valleys, wide lagoons, and offshore bars. Cap. 

The dead, blanched forms of up- 
lifted coral reefs also bear mute testimony to a relative uplift of certain 
tropical shores. 

Coasts frequently undergo uplift followed by submergence, or de- 
pression followed by later uplift. The result of such oscillations in the 
level of land and sea is the production of a variety of shore features 
some of which have resulted from emergence and others from submer- 
gence. Thus, for example, the Central Atlantic coast of the United 
States (Fig. 143) is marked by the drowned valleys characteristic of a 



the reef-building corals thrive best, for in the rush and dash of the 
waves they find the most food and calcium carbonate in the water, and 
the richest supply of life-giving oxygen. 

Coral reefs are sharply limited in their distribution by the depth and 
the temperature of the water. Although individual corals of various 
kinds live in more varied environments, the reef-forming kinds thrive 

Fig. 144. Growing corals, seen at low tide. The view is from a barrier reef across the 
lagoon toward mainland 20 miles away, Great Barrier Keef , Australia. (Saville Kent.) 

only where the water is clear, normally saKne, shallow, and warm. 
They cannot endure a temperature below 68 F. and they live normally 
where the water is less than 150 feet deep, though rarely, where there 
are descending warm currents, they exist well below this limit. Coral 
reefs are therefore confined to the shallow waters of the tropics and near 
tropics, except where (as at Bermuda) warm currents form exceptional 
conditions. The distribution of fossil coral reefs in the older geologic 
deposits is consequently thought to carry special climatic significance. 

According to their position and form, coral reefs have been grouped 
into three general classes: fringing reefs, barrier reefs, and atolls. 

Fringing reefs lie close against the shore and form a bench or platform 
extending out toward the sea, laid bare only at very low tide. As the 


corals cannot grow above sea level the growth of this platform is chiefly 


The width of the reef seems to depend largely on the steepness of the 
land slope. Since the corals grow and flourish chiefly on the outer edge 
of the reef, the seaward slope of the latter is usually very steep. As 
material is broken off by the waves and rolls down the slope, it forms a 
rising talus which graduaUy becomes compacted and cemented by 
calcium carbonate, eventually forming a base upon which the corals 
advance the reef seaward. If the bottom slope is steep, the debris 
settles into deep water and much upbuilding is required to give a sup- 
port for a small advance; if the slope is gentle the reef builds out more 
rapidly and may attain a width of a few miles. Opposite the mouths of 
streams the reefs are nearly always wanting because the corals cannot 
endure the freshened and muddy water. 

Barrier reefs differ from those just described in that they are situated 
some distance from the land, and are separated from the shore by a 
lagoon of shallow water. Many of the high volcanic islands of the Pacific 
as well as the islands of the Caribbean region are more or less completely 
girdled by such encircling reefs. The west coast of the island of New 
Caledonia has a reef of this sort that extends for 400 miles, and the 
greatest of all coral reefs is the Great Barrier Reef which stretches for 
1200 miles along the eastern coast of Australia, with an average dis- 
tance of 20 to 30 miles from shore and a depth of 100 to 300 feet of water 
in its great lagoon. 

A barrier reef may be from 1 to 30 miles from the shore and the aver- 
age of the maximum depths of the lagoons in the Pacific is about 200 
feet, though many of them are much shallower. Openings or breaks 
occur in the barrier because of the ebb and flow of the tide into the 
lagoons, and where they are sufficiently deep the lagoons serve as harbors. 

Atolls are more or less ring-shaped reefs enclosing circular lagoons 
instead of islands (Fig. 145). The breadth of such rings may be but a 
fraction of a mile or it may be as much as 20 or even 50 miles, and the 
depth of water in the lagoons ranges from a few feet up to 300 feet. 
On the outside the slope may descend quite sharply thousands of feet 
to the ocean floor. Generally there are openings in the reef on the lee- 
ward side, which afford access to the lagoon. 

Atolls, like either of the other types of reefs, may support low islands 
where storm waves have tossed the coral debris upon the reef and built 
it above sea level. The islands thus made are usually not more than 
10 or 15 feet above normal sea level and from a quarter to half a mile 
wide, though often they are long in the direction of the reef. Most of 
them are covered by vegetation, including palms, and form spots of 



great beauty. Some are inhabited, though their low elevation subjects 
them to the danger of being swept by the sea during exceptional storms. 
The only living atoll in America is small Sombrero of the West Indies. 


Fig. 145. An atoll. (After Dana, from an old picture.) 

The Origin of Barrier and Atoll Reefs. The origin of fringing reefs 
is easily understood. Since corals prefer shallow, warm water they tend 
to attach themselves near the shore and grow at first in scattered colonies 
which later coalesce as they increase in size. Thus a continuous coral 
growth is built up to low tide level, after which it can only grow outward 
as a fringing platform. It is equally clear that typical fringing, barrier, 
and atoll reefs are stages in a completely intergrading series. On the 
other hand the cause of the change from the fringing to the barrier or 
atoll form has given rise to much speculation, and the explanations 
proposed have invoked geologic changes of great significance. Three 
distinct theories may claim our attention. 

The subsidence theory, first advanced "by Charles Darwin, elaborated 
by Dana, and more recently rejuvenated by Davis, will be easily under- 
stood by reference to Fig. 146 (a), which represents a volcanic island that 
has slowly subsided. At an early stage, a fringing reef (A) has been 
formed. As explained above, the corals thrive best on the outward 
face of the reef. Accordingly, it may happen that after the reef has 
attained a considerable width only the more thrifty outer margin can 
grow fast enough to keep pace with the subsidence. The inner part 
then gradually becomes drowned and is covered by the lagoon behind 
the reef. As growth continues chiefly on the seaward face, the reef 



grows outward farther and farther. Portions of the coral skeletons 
broken by the waves fall down to make a talus outside the reef and some 
of the finer detritus is washed over the reef into the lagoon, which thus 
tends to be silted up gradually with limy mud as its bottom subsides. 
Stage B in our figure represents this phase in the island's history, with 
the barrier reef fully developed. As subsidence continues the volcanic 
island finally disappears below sea level, and the reef becomes an atoll 
as in stage C. It is evident that in an atoll formed in this way the bed- 

Fig. 146. (a) Sector diagram of a degrading and subsiding volcanic island encircled 
by coral reefs. Sector A shows the island at an early stage of its development when its 
reef is fringing; B shows the island partly submerged and deeply eroded and surrounded 
by a barrier reef; C shows a final stage when the island has been reduced to sea level and its 
reef has become an atoll. (Modified from Davis.) 

(6) Section through the island shown in (a) after it has become submerged and its reef 
has been transformed into an atoll. The successive levels of the sea correspond with the 
sea levels of sectors A, B, and C, respectively, in (a). Note how the fringing reef has 
been gradually transformed into a barrier reef and finally into an atoll through its upward 
growth as the island sank. 

rock of the original island will have a thick veneer of coral, as in Fig. 
146 (6). There follows as an important corollary of this hypothesis the 
belief that vast areas of the ocean now dotted with barrier reefs and 
atolls have undergone recent subsidence, in places amounting to thou- 
sands of feet. 
The solution theory, advanced by Murray and promulgated by Alex- 


ander Agassiz, sought to account for barriers and atolls without involving 
subsidence. It was for a time widely accepted. With a stationary sea 
level the reef grows outward because corals thrive best on its seaward 
margin. In the meantime, it is supposed by proponents of this theory, 
the dead portion of the inner margin of the reef, bored into by innu- 
merable organisms, crumbles and by the combined action of solution 
and of scour by the tidal currents is gradually removed, while in its 
place a lagoon appears. Thus a fringing reef becomes a barrier. It is 
also supposed that some of the existing barriers began their growth 
around the margins of submarine platforms where islands had been 
partly cut away by the waves. Atolls are supposed to have grown in a 
similar fashion where corals grew on shallow submarine banks, as on 
islands that had been completely truncated by the waves. As a corol- 
lary of this theory, the coral growth in barrier and atoll reefs should be a 
relatively thin veneer over a rocky platform. 

The glacial control theory, recently advanced by Daly, is in a sense a 
modification of the last. Daly calculates that the removal of marine 
water to form the continental ice caps of the last glacial period lowered 
the level of tropical seas from 200 to 250 feet. Owing to the colder 
condition of the Earth at that time (and thus of the sea), it is inferred 
that coral life was restricted to very narrow tropical belts. Elsewhere 
the oceanic islands, undefended by caps and belts of growing coral, were 
exposed to the erosive action of the waves, which cut wide terraces 
around the harder and larger islands and cut off those that were small or 
were formed of softer, less compact material. When the ice melted, 
and the seas grew warmer, the corals returned to these terraced or 
truncated islands, growing best at the seaward margins of the wave-cut 
platforms where they built up new reefs. Meanwhile the inner parts 
of the wave-cut platforms were drowned by the rising sea level as the 
glacial ice melted away, developing into lagoons. Daly sought in this 
way to account for the fact that most of the larger lagoons now have a 
depth of 200 to 250 feet and that the majority of the reef-fringed coasts 
show evidences of recent drowning. 

In conclusion, it should be noted that these theories are not necessarily 
exclusive of one another. One process may have operated in some lo- 
calities and another elsewhere. Nevertheless, recent chemical investi- 
gation has shown that the water in the lagoons is saturated with calcium 
carbonate and instead of exerting a solvent action tends to precipitate 
calcium carbonate. Furthermore, Davis 7 extensive study of the oceanic 
islands surrounded by barrier reefs shows that they very generally have 
embayed shorelines due to the drowning of their lower valleys. This 
feature proves clearly a relative sinking of the islands, though whether 


it has been an actual subsidence of the land or a rise of sea level must be 
inferred from a study of the form of the barrier-encircled islands. 

An island maturely dissected by streams and then slowly submerged, 
while protected from the waves by a growing reef, presents a shoreline 
marked by open embayments separated by sloping and non-cliffed 
spurs; but the formation of a deep wave-cut terrace at a time when coral 
growth had ceased must have developed a sea-cliffed shoreline, and the 
subsequent drowning by a rise of sea level should leave open embay- 
ments between steeply cliffed spurs. Davis (1928) has shown that while 
many of the islands near the margins of the tropical coral reef zone have 
cliffed spurs, those in the main coral reef regions ordinarily do not. 
The conclusion follows that glacial control has played a part in the 
outer areas of the tropics where reduced temperatures would first in- 
hibit the growth of the corals, whereas in the great coral reef areas in 
the Pacific, subsidence must be the chief cause of the formation of 
barrier and atoll reefs. Recent verification of such extensive subsidence 
in the Marquesas Islands group has come from a study of the distribution 
of the flora of the islands, which indicates a recent subsidence of from 
3000 to 5000 feet. 

Finely divided calcareous sediments are now accumulating exten- 
sively on warm shallow submarine banks like that bordering southern 
Florida and the Bahama Islands where little terrigenous sediment is 
supplied to the sea. The conditions under which these deposits are 
being formed are discussed on pages 222-223. 

Oozes of the Ocean Bottom. The abyssal ocean floor, constituting 
three-quarters of the Earth's surface, is covered to an unknown depth 
by exceedingly fine-grained, soft deposits known as oozes. These sedi- 
ments are of several kinds and of different sources. 

Red clay is the most extensive of the oozes since it mantles the deepest 
parts of the ocean floor to an extent of over 50,000,000 square miles, an 
area about equal to the entire land surface of the Earth. Its sources 
are believed to be windblown dust (derived both from erosion of the 
land and from volcanic eruptions), pumice that has floated for a time 
before sinking, meteors and meteoric dust that have fallen directly into 
the oceans, and also the insoluble residues of organic structures that have 
sunk to the ocean bottom. All of this material is so thoroughly altered 
by chemical decay that the original source material for any sample of 
the clay is obscure. The red color of this extensive deposit is probably 
the result of the extreme slowness of its accumulation, which gives time 
for its thorough oxidation in spite of the deep covering of marine water. 

The remaining abyssal deposits are the organic oozes formed largely of 
minute shells or shell fragments dropped by organisms that live in the 


surface waters of the ocean. Some of these deposits are calcareous, 
being made of limy shells, and others are siliceous, being formed by the 
accumulation of microscopic siliceous shells. 

Several distinct types of organic ooze are recognized, each being 
named for the group of organisms most important in its formation. 
The most widespread of the limy oozes is that known as globigerina 
ooze (Fig. 147). It is formed largely of the microscopic shells of single- 
celled animals known as Foraminifera. There are about twenty species 
of these animals that live floating in 
incredible numbers at the surface of 
the warmer oceans. In the repro- 
duction of these tiny creatures the 
mother forsakes her shell and subdi- 
vides into a great many tiny daugh- 
ters, each of which in turn secretes 
a new shell as it grows up, only to 
repeat the process. As reproduction 
is rapid and generation succeeds K& 147. - GioHg^ ? e muc \ h 

. . magnified. (After Agassiz and Murray.) 

generation quickly in such forms, 

their abandoned shells drift down like a perpetual snowfall into the 
depths. Those of the genus Globigerina usually predominate, hence the 
name given to the deposits. Wherever these little shells fall below a 
depth of about 15,000 feet, the carbonic acid concentrated in the cold 
bottom waters again dissolves the shells, so that globigerina ooze is not 
found in the deeps of the ocean, but on the shallower parts of the abyss it 
mantles a vast expanse of approximately 30,000,000 square miles. 

Radiolarian ooze is formed by the Radiolaria, another group of floating 
microscopic animals which, unlike the Foraminifera, fashion their delicate 
and beautifully ornate shells from silica. Diatom ooze is likewise formed 
by a group of microscopic plants, the diatoms, which secrete capsule-like 
shells of silica. 

The oozes of the abyssal region are distinctive, and the general rarity 
of this type of deposit in the sedimentary rocks of the lands is believed 
to indicate that the continents have always remained continents and with 
very limited exceptions have never been covered by abyssal depths of 
water. Locally, rocks formed from abyssal oozes are present above 
sea level, as in the Island of Barbados, in some of the Netherlands East 
Indies, and possibly the Apennines and the southern Alps (Steinmann), 
but their total area probably is not great. 



1. Shore Processes and Shoreline Development; by D. W. Johnson. 584 pages. 
John Wiley & Sons, New York, 1917. 

An exhaustive discussion of the physiographic development of the shore region. 

2. New England-Acadian Shoreline; by D. W. Johnson. 608 pages. John Wiley 
& Sons, New York, 1925. 

3. The Depths of the Ocean; by Sir John Murray and J. Hjort. 821 pages. 
The Macmillan Co., 1912. 

An extensive account of the methods of oceanographic investigation and of oceanic 
currents, temperature and depths as well as a description of oceanic life and bottom 

4. An Introduction to Oceanography; by J. Johnstone. 368 pages. University 
Press of Liverpool, 2nd edition, 1928. 


Sedimentation and Stratification. The muddy water of swollen 
streams is turbid with sediment in transit toward the sea. Some of this 
material comes to rest in lowland areas where the streams spread in flood, 
but the final goal for most of it is the sea floor. In either case the sedi- 
ment is spread in layers of mud or sand, and much of it is compressed and 
cemented eventually into sedimentary rocks. 

The layered or stratified nature of such deposits arises from an in- 
termittent supply ctf sediment or from changes in the velocity of the 
currents. If, for example, sediments of various degrees of fineness were 
dropped together into still water, the heaviest and coarsest would reach 
bottom first and upon them would settle the next in size and so on up 
to the very finest. There would, of course, be no distinct layers but a 
complete gradation from bottom to top. But if a second lot of sediment 
were introduced after the first had settled, it would form a distinct layer 
with its coarse base resting on the fine-grained portion of the first layer. 
In similar fashion the sediments introduced into a lake or the sea after 
one storm will spread and settle before the next storm or flood. If, 
on the other hand, a mass of heterogeneous sediment is dropped into a 
steady current of moving water, the gradation takes place in a horizontal 
as well as a vertical direction, the successively finer material being 
dropped farther and farther along the bottom, or, in other words, 
being graded but not stratified. If the velocity of the current is altered 
and a second supply of sediment introduced, the new deposit will be 
evenly graded like the first, but its coarseness at any point will not 
correspond exactly with that of the first layer immediately below it. 
The two layers will be separated therefore by a distinct bedding plane ,* 
on either side of which they differ in texture. Since the velocity of 
streams and of the shallow marine currents changes from day to day 
and varies from place to place, it is inevitable that the sediments they 
transport will be deposited in parallel layers that differ in thickness, 
texture, and materials. Each single layer is known as a stratum, the 
bedded or layered deposits are said to be stratifiedj and the condition is 
known as stratification (Fig. 148). 




Source and Nature of Sediments. It has been shown in previous 
chapters how each of the erosive processes contributes its share to the 
endless stream of rock waste which the rivers move toward the lowlands 
and the sea. This land waste varies in coarseness from the great 
boulders moved by glaciers or torrential streams to the finest of mud; 
and it also includes material which, like salt and calcium carbonate, is 
borne in solution. At the same time, it varies in composition from bits 
of fresh and unweathered rock to the end products of chemical decay. 

Tig. 148. Regularly bedded sandstones and shales, near Pueblo, Colo. (U. S. Geol. 


It is roughly classified, chiefly on the basis of coarseness, into gravel, 
sand y silt, and day. 

Gravel is an aggregate of coarse sediment in which individual particles 
have a diameter of 2 mm. or more. The smaller pieces, ranging in size 
from that of grape seed to that of baseballs, are known as pebbles, and 
the larger ones as boulders. Besides the pebbles and boulders, gravels 
usually include more or less sand. 

The fragments that make up the gravel are always more or less water 
worn. At their source they were irregular, angular pieces of rock 
bounded by joints or fracture planes, but in being rolled along down- 
stream, or tossed back and forth by the waves, their angles and corners 
suffer the most abrasion and the pebbles and boulders become more and 
more perfectly rounded. Rough and subangular fragments, therefore, 
indicate that they have suffered little transportation and that they do 
not lie far from their place of origin. 


In their transportation, the pebbles and boulders of softer material 
are the first to be destroyed by reduction to sand and gravel. Since 
quartz is the hardest of the common rock-forming minerals it is the 
predominant substance in most gravels; and well rounded pebbles and 
boulders, having suffered long wear, are usually composed almost 
entirely of quartz. On the other hand, coarse sediment that has not 
travelled far commonly includes other minerals and rocks, such as 
granite, schist, basalt, or limestone. 

Sands are sediments composed of grains smaller than gravel, yet 
exceeding 1/16 mm. (i.e., about the thickness of this page) in diameter. 
Such material is coarse enough so that it will not form a coherent, plastic 
mass when wet. Ordinary sand is like granulated sugar in fineness. 
With a lens it may be seen that the coarse sand grains are rounded like 
pebbles by attrition, but the finest grains are usually angular. This 
does not mean that the larger grains have travelled farthest but rather 
that grains below a certain size are not rounded in transit, the explana- 
tion lying in the fact that the film of water between them serves as an 
effective buffer to protect such tiny particles against abrasion. 

Quartz so commonly forms sand that, unless otherwise stated, quartz 
sand is generally understood. Many other minerals occur in sands, 
however, and the beaches of coral islands are in places formed of " coral 
sand " made wholly of calcium carbonate. 

Silt and clay include the finest part of the land waste. This finely 
divided material coheres when wet, and when dry will crumble into dust. 
The distinction between silt and clay is largely one of fineness of sub- 
division, clay consisting of microscopic particles (less than 1/256 mm. 
in diameter) and silt of somewhat coarser sediment. Due to its most 
finely divided (colloidal) constituents, clay is plastic when wet whereas 
silt tends to crumble as would very fine sand. The common term mud 
is applied to either silts or clays and is used particularly where the sepa- 
ration of silt and clay is very imperfect. 

The mineralogical composition of silts and clays is varied and much 
more complex than that of sand. The most characteristic constituents 
are hydrous silicates of aluminum and the alkaline earths and hydrous 
oxides of iron. Some silts, like those derived from glaciers, are produced 
by rock grinding instead of by chemical decay and these deposits may 
show a predominance of unweathered particles of all sorts of rocks and 

The materials carried in solution form a large but invisible contribu- 
tion to the sea, reappearing as sediments where evaporation or chemical 
reactions cause their precipitation, or where organisms extract the dis- 
solved minerals to form shells. 


Volcanic ash, wind-blown dust, glacial till, and organic materials such 
as peat are other deposited materials that enter into the formation of 
special types of sedimentary rocks, but although of interest and local 
importance, they "are not of such geological significance as those men- 
tioned above. 


In Chapter IV it was explained how streams transport their burden of 
sediments. The spreading of marine sediment by the waves and cur- 
rents has been discussed in Chapter VIII. The conditions under which 
the sediments finally come to rest must now claim our attention. Three 
major realms of deposition are generally recognized. The first is that 
of the land areas. The deposits formed there are known as continental 
sediments since they belong essentially to the continents, whether formed 
actually on the dry land or under the surface of its streams and lakes. 
The second realm is that of the intertidal zone and its deposits are known 
as littoral sediments. The third is that of the marine realm and its de- 
posits, formed anywhere over the sea or ocean floors, constitute marine 
sediments. The distinction between these three realms is of such geologic 
significance that each of them deserves more detailed consideration. 

Regions of Continental Deposits. Although vast areas of the land 
surface are undergoing active erosion, about 10,000,000 square miles 
of the continents are now buried by accumulating sediments or by sedi- 
mentary deposits formed in the recent geologic past. These regions of 
deposition may be classified into four types in which the conditions of 
accumulation are strikingly different. These are (1) piedmont plains, 
(2) arid basins, (3) humid basins, and (4) great deltas. 

Piedmont alluvial deposits are formed where streams debouch from 
relatively young and lofty mountain ranges. Here the velocity of the 
streams is rapidly checked as they enter the flatter piedmont belt and 
they drop a part or all of their load. During times of flood, when on 
leaving the mountains they are heavily laden with sediment, the streams 
spread widely beyond their channels and cover the adjacent country with 
layers of sand and clay. In so doing, they overflow into the lowest 
places, filling them with sediment, and by the continued action of this 
process through a long period of time they build up deposits of great 
thickness in the form of alluvial plains in front of the mountains. Strik- 
ing examples are to be seen in the deposits that underlie the High 
Plains east of the Rocky Mountains, the Pampas east of the Andes, 
and the Indo-Gangetic Plain south of the Himalayas in India. 

Since they are formed on the flanks of the mountains and well above 
sea level, these sediments will ultimately be destroyed as the erosion 


cycle progresses, unless preserved by deep down warping of the piedmont 
region. Even now, for example, the very streams that formed the 
High Plains are stripping these deposits away again. In this case 
the stripping may be hastened by climatic changes rather than by 
reduction of the highlands. 

Desert deposits occupy great areas in the arid basins. Almost one- 
tenth of all the land surface of the world is embraced in interior basins, 
whose drainage is centripetal, with no outlet to the sea. Here the waste 
of the slopes constantly tends to move toward the deeper parts of the 
basins and accumulate in stratified deposits. At rare times of heavy 
rainfall the temporary streams spread widely over the lower slopes, 
shifting sand and gravel toward the center of the basins, while the finer 
material is swept on into temporary lakes (playa lakes) that form in the 
deeper depressions. Permanent lakes like Great Salt Lake or the Cas- 
pian Sea tend to be filled by the sediments of the larger and more per- 
manent streams. Moreover, in times of dryness, sand is shifted by the 
wind. Except for wind-blown dust none of the sediment can escape 
and thus, through the continued action of rain wash and streams, the 
desert basins tend to fill with deposits. If sinking of the basins occurs 
while they are being silted up, the ultimate thickness of the sedimentary 
deposits thus formed may attain several thousands of feet. Deposits 
of arid basins commonly contain layers of salt and gypsum as well as 
thick beds of cleanly sorted dune sands. They are generally poor in 
fossils and exhibit light colors such as dun, buff, or pink. Some of them 
are red. 

Deposits in humid basins also may be of vast extent. Where struc- 
tural basins form rapidly in regions of abundant rainfall, great lakes 
are produced; but if the sinking is slow, filling may keep pace with sub- 
sidence and the region remain a swampy lowland. The second alter- 
native is favored where the basin has bordering highlands undergoing 
rapid erosion, for then the incoming streams are heavily laden with 
sediment. A good example of this sort exists in the upper basin of the 
Paraguay River in South America where an area about 400 miles long 
and as much as 150 miles wide remains a labyrinth of lakes and swamps 
and channels in a low grassy plain. During the annual rainy season the 
whole area is flooded, but during the rest of the year only about one- 
fourth of it is covered. The river enters the lowland heavily burdened 
with sediments brought by tributaries from the Andes to the west and 
from the Brazilian highland to the "north. The water leaves the basin 
fairly clear, having spread most of its sediments over the lowland to be 
mingled with organic matter from the decay of vegetation that flourishes 
abundantly in such places. This swampy condition, with resulting 


accumulation of river sediment, may be indefinitely maintained if the 
basin is a region of continued subsidence. Though deposited in water, 
such sediments are to be regarded as continental in origin, since they 
occur in hollows of the land surfaces. Such deposits have accumulated 
extensively in times past and seem to represent approximately the con- 
ditions under which the older coal-bearing rocks have been formed. 
The abundant organic matter gives such sediments somber or dark 

Deposits formed in this manner are likely to escape destruction by later 
erosion because they lie only slightly above sea level. 

A delta is a deposit of sediment built out by a stream into standing 
water. An idealized vertical section of a delta (Fig. 149) shows that it is 
partly above and partly below water level. Even the landward portion 


Fig. 149. Section through a delta built into quiet water of constant level. A, bottom- 
set beds; B, foreset beds; C, topset beds. (After Barrell.) 

is extensively covered by fresh water when the stream is in flood and 
carrying its greatest load of sediment. At such times sand and silt are 
spread far and wide by the flood waters, forming nearly horizontal layers 
over the landward top of the delta. The sediment that reaches the shore 
is shifted by the waves and currents toward the sea, building out the 
shallow submarine surface like a vast embankment with a nearly 
horizontal surface near shore, giving way to a steeper slope beyond the 
line of most rapid building. The sediments that come to rest on the 
nearly horizontal upper surface of the delta, whether the landward or the 
subaqueous portion, are known as the topset beds. The material carried 
out over the inclined front of the delta constitutes he foreset beds, 
and that which is spread still farther, parallel to the sea bottom, forms 
the bottomset beds. The topset beds are partly continental but the foreset 
and bottomset beds are marine. 

As shown in Fig. 149, when the level of sea is stationary the delta growth 
is chiefly seaward by an extension of the foreset beds, and in this case 
the continental sediments, notwithstanding their areal extent, form but 
a small part of the volume of the delta. When a delta is subsiding, on 
the contrary, more of the sediment comes to rest over its upper surface 
where the sluggish waters spread in flood. If the area of the landward 



portion of the delta is great, it may thus happen that the volume of 
land deposits, formed by the topset beds, is vastly greater than that of 
the f oreset beds which build out the delta front (Fig. 150) . Consequently, 
a delta which is built upon a subsiding foundation tends to form dom- 
inant topset beds of continental nature. The deltas of great rivers, like 
those of the Mississippi and the Nile, are built out into shallow seas on 
the wave-swept continental platforms, and here the difference in the 

Fig. 150. Section through a delta built into quiet water but resting on a subsiding 
foundation. The unshaded basal part of the delta is assumed to have formed before 
subsidence began. Since subsidence of the delta started, the shoreline has advanced 
landward and much of the sediment has come to rest as topset beds which are partly sub- 
aerial and partly submarine in origin. (After Barrell.) 

inclination of foreset and bottomset beds is slight and the distinction 
between these two parts of the delta j.s not marked. 

Regions of Littoral Deposits. The total area of the present littoral 
zone in all the continents is probably not much over 60,000 square miles. 
On ordinary shores it is usually far less than 2 miles wide, though in 
great deltas like that of the Mississippi the salt-water marsh may be 
25 miles or more in width and littoral sediments may accumulate rapidly 
thereon. The most remarkable extension of the existing littoral zone 
occurs in the so-called Runn of Cutch, a low flat on the southeast side 
of the Indus River delta that has an area of about 6000 square miles. 
It is so flat and so nearly at sea level as to be flooded with marine water 
during one season of the year when monsoon winds blow inland, and laid 
bare or covered by overflow from the Indus when the winds blow off- 
shore. Some of the older mud-cracked marine formations may have 
formed under comparable conditions. In general, however, littoral 
/deposits cannot be of great thickness, for, if the land is building out into 
f the sea, they must give place to land deposits and be buried under them; 
if the sea is encroaching on the land, they must yield to marine sedi- 
ments and be covered by them. In either case, they stand a rather 
poor chance of permanent preservation, since upon slight uplift they 
are exposed to stream erosion and with stationary sea level the waves 
gradually encroach upon the shore zone and remove them. They are 
likely to be preserved only in subsiding large deltas. 


Regions of Marine Deposits. The greater part of the land waste 
finally comes to rest on the sea floor. Most of this terrigenous (land- 
derived) sediment accumulates in the neritic zone but the finest muds 
can be transported even beyond the edge of the continental shelf and 
are found as much as 200 miles from land. Coarse materials accumulate 
in shallow water near the shore whence they are supplied, for there alone 
are the currents strong enough to distribute gravel and coarse sand. . 
Fine sands extend in places to a depth of 200 or 300 feet or even more and 
beyond them the terrigenous muds spread as a continuous blanket not 
only to the continental margin but part way at least over the continental 
slope, where they grade imperceptibly into the oozes of the ocean floor. 
It is improbable, however, that much land-derived mud reaches the 
abyssal region. While in general there is a seaward gradation from coarse 
to fine sediment, it must not be supposed that all the muds are deposited 
in deep water, nor that there is always a shoreward phase of sandy 
sediment between the land and the place where mud deposition begins. 
In many protected embayments, as the Baltic Sea or Chesapeake Bay, 
muds, even of the finest sort, are now being deposited right up to the 
shore line. 

Calcareous deposits of the neritic zone form chiefly in shallow water 
where clastic (fragmental) sediments are scarce or absent and where 
lime-secreting organisms grow luxuriantly. At the present time limy 
deposits are forming chiefly about the tropical coral islands and over 
warm shallow submarine banks such as that bordering Florida on the 
south and the Bahama Islands on the west. It has sometimes been 
wrongly inferred that marine sediments grade in depth from gravel or 
sand through muds to limy ooze, and it should be emphasized, therefore, 
that the limestones now exposed on the continents represent deposits 
in shallow, not deep water. 

The vast deposits of red clay and organic oozes that cover the deep 
ocean floor have been described in the previous chapter. They are of 
enormous extent, but since such deposits are essentially lost to the conti- 
nents and only rarely have been elevated into lands their geologic in- 
terest is limited. 


Stratified rocks are merely sediments consolidated into stone. Com- 
monly the appearance of the rock clearly suggests this origin, but some 
of the older rocks have been modified so greatly that the resemblance 
to sediments would be obscure if it were not for their stratification and 
entombed fossils. The consolidation is the result of several factors, 
chief of which are pressure and cementation. As the sediments accumu- 



late, the upper layers bear upon all those buried below. If the specific 
gravity of loose sediment is 2.3, the pressure will increase by fully 1 
pound per square inch for each foot of depth. At a depth of 1000 feet, 
therefore, the weight of the overlying deposits is at least 1000 pounds 
per square inch. As a result of such compression, sediment is made more 
compact as the particles are squeezed tightly together and most of the 
water is pressed out. 

Cementation results from the deposition of mineral matter in the spaces 
between the grains of sediment. The most common cementing sub- 
stances are calcium carbonate, 
silica, and iron oxide. These 
cements may be introduced in so- 
lution in the water that fills the 
pore space when the sediments are 
deposited, or they may be intro- 
duced later by percolating ground 
water. Some sandstones have 
but little cement and therefore 
crumble readily into loose sand 
when exposed to the weather, but 
others have the voids almost com- 
pletely filled with cement (Fig. 
151) and have been thus con- 

verted into firm rock. 

. . 
The interior heat OI the lLarth 

risine; into SUCh masses Of Sedi- 

ments may aid in some degree to 
consolidate them by quickening the chemical activity of the diffused 
waters which deposit the cement. And finally, since the conversion of 
sediments into rock must be a slow process, time is an important element. 
Thus in general the more recent sediments, where they have been ex- 
posed, are softer and more friable than the older ones. It must not be 
inferred, however, that the process of cementation takes place only 
under the sea, for on land, solution by ground water, transfer to lower 
levels, and redeposition of cementing materials take place on a large 


The different kinds of sedimentary rocks depend mainly upon the 
nature of the sediments from which they are formed. The chief types 
of sediment and the equivalent sedimentary rocks are as follows: 

Fi - 151 - Microscopic section of a firmly 
cemented sandstone. The dotted areas are the 
rounded sand grains and the clear areas repre- 
^ silica de P osited between the grains, bind- 
ing them into rock. 



Compacted Strata, as Rocks 






It must not be imagined that the different kinds of rocks mentioned 
above are always sharply defined from one another as wholly distinct 
types. Just as muds grade through sand into gravel, and pure cal- 
careous deposits into silts or clays, so may the various rocks formed from 
them grade into one another. In the description of these impure or 
mixed types of rocks, those that were muddy or clayey sediments are 
termed argillaceous (from the Greek argillos, clay). Similarly, sandy 
ones are described as arenaceous (from the Latin arena, sand) and limy 
sediments or rocks are described as calcareous (from the Latin calx, 
lime). These three adjectives are used in the discussion either of 
sediments or of the rocks derived from them. 

The chief types of sedimentary rocks deserve more detailed discussion. 

Conglomerate. A typical conglomerate is shown in Fig. 152. It 
consists of rounded pebbles set in a matrix of well-cemented sand. 
Generally the pebbles and boulders in conglomerates are made of 
quartz, but they may consist of fragments of any kind of rock. The 
fragments vary in size from a small fraction of an inch to many feet in 
diameter. They may be packed together with little fine matrix or 
mingled with any proportion of fine material. If the pebbles are few 
and scattered through a sandy matrix, however, the rock should be 
called a pebbled sandstone rather than a conglomerate. 

The pebbles and boulders may be well rounded or, if they have suffered 
but little transportation, they may be more or less angular. If they 
are distinctly angular the rock is a variety of conglomerate known as 

A special type of conglomerate or breccia is produced when the layers 
of accumulating sediments,, indurated before complete burial, are broken 
up and the dislocated fragments rolled about and then recemented. 
This happens where exceptional storm waves violently stir up the 
bottom, where layers exposed to the atmosphere are broken up by 
mud cracks, where streams in their floods undercut the sides of their 
channels, or where mountain-making disturbances rupture the beds on 
the sea floor. The conglomerates thus formed represent merely an 
episode in the deposition of the formation in which they lie and they are 
therefore known as intraformational conglomerates. They are made of 


the same materials as the enclosing beds and do not signify uplift and 
renewed erosion as do the other types of conglomerates formed of im- 
ported gravel. Usually the individual fragments of the intraformational 
conglomerate are only imperfectly rounded and they may be quite 

Arkose is a special variety of conglomerate or sandstone containing 
much feldspar. Its occurrence indicates that the component material 
was not long exposed to weathering before it was deposited, and therefore 
probably was not transported great distances. Arkose is more corn- 

Fig. 152. A piece of conglomerate, shown about half natural size. 

monly of continental than marine origin, and its formation is favored 
by the breaking down of granitic rocks in cold or arid climates where 
rock decay is inhibited. 

Sandstone. Sandstones are composed of cemented sand grains, 
which consist most commonly of quartz. Many sandstones are quite 
even in grain, but there are hybrid varieties grading on the one hand 
into conglomerate, and on the other into shales. In red and brown 
sandstones the cement is mainly oxide of iron but in white, buff, or gray 
varieties it is most commonly either silica or calcium carbonate. Sand- 
stones are generally porous and the interspaces may amount to 30 per 


cent of the total volume. For this reason they are favorable reservoirs 
for artesian water and petroleum. 

Shale. The indurated equivalent of silts and clays is shale. The 
majority of shales possess a more or less thinly laminated structure 
resulting from closely spaced, parallel bedding planes. Such shales tend 
to split easily along these natural planes of stratification. If the lami- 
nae are thin and flexible the shale is said to be fissile. On the other 
hand, some shales occur in layers of considerable thickness that show no 
subdivision by bedding planes and tend to break into blocks instead of 
laminae. Such massive or blocky shales are sometimes designated 

Shales are soft and generally weak rocks, crumbling readily into small 
chips. They show a great variety of colors as do the muds from which 
they form. Unlike sandstone, they tend to be impermeable to water. 
Since they are composed to a greater or less degree of clay they yield 
a strong and characteristic clay odor. 

Shales occur in both continental and marine deposits. In the former, 
they represent mostly the flood-plain deposits well back from the chan- 
nel where roily waters, impounded as backwater or at least retarded in 
their flow, have been unable to keep the fine sediments in suspension. 

Black shales constitute a distinctive type of sedimentary rock com- 
monly associated with coal beds and also occurring abundantly in marine 
deposits, where they are important as source rocks in the geological 
formation of -petroleum. The dark color of these shales is due to 
carbon derived from the incomplete destruction of organic matter buried 
with the mud. An abundant supply of organic matter and stagnation 
of the water appear to be the essential requirements for black shale 
formation, the stagnation leading to a deficient supply of dissolved oxy- 
gen so that decay is inhibited. Depth is not a vital factor, though 
obviously any sharply depressed areas or hollows in the sea floor are 
likely to be passed over by bottom currents and to become stagnant 
black mud holes. Black muds are now accumulating in both shallow 
and deep water. Where black muds are forming, hydrogen sulphide 
(H 2 S) is usually generated and it in turn reacts with iron salts in the sea 
water to form pyrite (FeS 2 ), which is precipitated in the form of con- 
cretions in the black shale, or very commonly in the form of replace- 
ments of fossil structures. 

Limestone. Limestone is the consolidated equivalent of limy ooze, 
calcareous sand, or shell fragments. It is composed of the mineral 
calcite or calcium carbonate (CaC0 3 ). Pure varieties of limestone are 
white or light gray, but impure varieties are buff, brown, red, dark gray, 
or even black. If fine-grained and nearly pure, the rock is compact 


and tough and forms one of the strongest of structural stones; but with 
an admixture of clay, limestones grade into calcareous shale which weath- 
ers readily into clay. Some limestones have a dense texture, others are 
finely crystalline, and still others distinctly granular. Some impure 
varieties appear earthy. 

Chalk is a soft, porous variety of limestone. Some chalks are com- 
posed largely of the microscopic shells of Foraminifera, the tiny animals 
that also make the globigerina ooze of the ocean floor. These minute 
shells are extremely fragile and commonly occur much broken. In 
addition to the shell fragments there is normally a matrix of fine par- 
ticles of calcium carbonate that has been considered to be a chemical 
precipitate. Some chalky deposits, as those of Kansas and Alabama, 
are made up largely of such finely divided material, with an admixture 
of clay, and show but few of the foraminiferal shells. 

Chalk was once supposed to represent deep-water deposits like the 
modern oceanic oozes; but in spite of the fact that globigerina ooze will 
probably form chalky deposits, it is now known that the great chalk 
beds of the present lands include shells of shallow water animals and are 
associated with coarse-grained sediments that could not have been 
washed into deep water. 

Coquina is a variety of limestone made up of shells and coarse shell 
fragments heaped together and loosely cemented. 

Dolomite is a rock formed largely of the magnesium-calcium carbonate 
mineral, dolomite, CaMg(C0 3 ) 2 . It may be considered a special 
variety of limestone. In fact, there are probably all gradations between 
calcite and dolomite limestones, and the mixed types are more common 
than pure dolomite. Dolomite resembles calcite limestone in most of 
its characters and is commonly regarded as limestone by others than 
specialists. It is of great extent and importance in the older rocks, 
especially. The Dolomite Alps of the eastern Tyrol, for example, have 
been carved out of dolomite formations 3000 to 4000 feet thick, and 
equally great masses of this kind of rock occur in Alabama and in western 

Limestone represents the calcium carbonate that is carried in solution 
to the sea and there precipitated; but the conditions for its deposition 
are less obvious than those under which the clastic (fragmental) sedi- 
ments are formed and there is still much to learn about the relative im- 
portance of the different agents that precipitate the limy sediments. 

The soluble bicarbonate of calcium is very unstable and can easily be 
made to give up one molecule of its CC>2 according to the formula 
H 2 Ca(C0 3 ) 2 = C0 2 + H 2 + CaC0 3 ; but when this happens the 
corresponding molecule of CaCOs becomes insoluble and is precipitated. 



Recent investigations have shown that the warm shallow marine water 
is essentially saturated with calcium carbonate whereas the colder, deeper 
water is undersaturated because of its richness in C0 2 . Wherever the 
deeper water rises in ocean currents to flow over shallow submarine 
banks in tropical regions the rise in its temperature tends to drive off 
part of the C0 2 and to leave the warmed water supersaturated with 
CaC0 3 . Chemical precipitation would occur in such places if the lime 
were not rapidly extracted by organisms before the saturation point is 
attained. However, these places afford the most favorable conditions 
for animals and plants that are prodigal in their use of carbonates to form 

Fig. 153. Mudflats off the west coast of Andros Island at low tide, showing the 
nature of the limy deposits that cover large areas of the Bahama Banks. The sediment is 
a soft white paste of nearly pure calcium carbonate. (R. M. Field.) 

shells or skeletal structures. Corals thrive and form reefs on such 
tropical shoals and with the corals are associated lime-secreting algae 
and a host of shell-forming creatures. 

Impressed by the limy mud and coral sand formed about coral reefs, 
the earlier naturalists attributed to corals a predominant role in the 
precipitation of the limy deposits. Many limestones, however, present 
little evidence of a coralline origin and recent investigations have shown 
that modern calcareous sediments of great extent are being precipitated 
directly as a fine ooze or mud. One of the most extensive areas of 
modern limy sediment is that of the Bahama Banks south of the Florida 
Straits. The total area of this great shoal is over 7000 square miles and 
the average depth of its water is less than 20 feet. The bank is formed 


of limestone and great areas of its surface are mantled by fine white 
limy ooze (Fig. 153). The agent of deposition of this limy material is 
still uncertain. Coral reefs are limited to the extreme margins of the 
shoal and appear not to be large contributors to the deposit at the pres- 
ent time. It has been thought that a bacterium (Pseudomonas calcis) 
caused the precipitation of the calcium carbonate but doubt has recently 
been cast on this idea. The relative importance of chemical and bio- 
logical agents in the deposition of the carbonate sediments is therefore 
still unknown. 

Diagenesis of the Calcareous Sediments. Important chemical 
alterations not uncommonly take place in the limy sediments during 
or shortly after their deposition. For example, no dolomite is formed 
directly, either by chemical precipitation or by organic secretion. The 
deposit goes down first as calcite and then, while still in contact with 
marine water, some of the calcite is replaced by magnesium. 

Another type of alteration is to be seen in the solution and recrystal- 
lization of the fine particles of limy mud to form granular or coarsely 
crystalline limestones, or, about coral reefs, to transform the loose coral 
debris into compact stone. 

All such chemical alterations as modify the sediments wnile tney are 
accumulating are embraced under the term diagenesis (Gr. dia, in two 
parts, + genesis, birth). 

Minor Sedimentary Rocks. Coal, iron ore, rock salty gypsum, 
and chert or flint are sedimentary rocks of special interest, but from the 
geologic point of view they occur in volumes so limited in comparison 
with the enormous bulk of the shales, sandstones, and limestones that 
they are of little importance considered merely as rock masses. 

Coal and iron ore are given special consideration in Part II of this 

In addition to the stratification which these rocks always exhibit, 
other features are commonly displayed which throw light upon the 
geologic history of the rocks. Among these features are fossils, mud 
cracks, ripple marks, and cross-bedding. 

Fossils. Fossils (Latin fossileSj from fodere, to dig up) are the re- 
mains or imprints of animals or plants that were buried with the ac- 
cumulating sediments (Fig. 154). The organisms are rarely preserved 
entire; generally only the hard structures, such as shells and bones, en- 
dure. These parts may be preserved without change, but commonly 
the original material is replaced, one molecule at a time, by mineral 
matter so that the fossil becomes a solid stony object or petrifaction. 


Not uncommonly the organic structure is wholly removed, leaving only 
an imprint or hollow mold in the rock. Natural molds of this sort are 
not infrequently filled by subsequent mineral deposits which then form 
natural casts of the organic objects. Finally, footprints or trails 
made by animals crossing soft mud are preserved under favorable 

Since most fossils represent the animals and plants that were living 
where the sediments accumulated, they throw much light on the condi- 

Fig. 154. Fossil shells in stone. 

tions that prevailed during the deposition of the f ossiliferous rocks. They 
show, for example, whether the beds were laid down on the land or be- 
neath the sea, since continental sediments contain the remains of land 
plants, bones of land animals, or shells of river clams, whereas marine 
sediments contain representative marine shells. Something of the cli- 
mate of the region at the time of deposition may be indicated also; for 
example, fossils of tropical plants and animals, such as palms and alli- 
gators, are found associated in certain formations of the " Badlands " 
of South Dakota. The manner in which fossils are used in deciphering 
the past history of the Earth and its inhabitants is considered in detail 
in the second part of this book. 



Mud Cracks. Soft muddy sediments left exposed after the recession 
of flood waters shrink and crack into characteristic polygonal blocks 
like those shown in Fig. 155. These desiccation fractures are known as 
mud cracks. Further exposure to air and sun bakes and hardens the 
blocks of mud. During the dry season wind-blown sand or silt may 
cover the surface, filling the cracks with sediment coarser than the mud- 
cracked layer. In this way the form of the polygonal blocks is preserved 

Fig. 155. Modern mud cracks formed on the delta of the Colorado River. 

Geol. Surv.) 

(IT. S. 

even after the return of later flood waters. After the whole series of 
deposits has subsided and become hardened into rock, the layers of 
shale and sandstone may later be exposed, exhibiting these " fossil " 
mud cracks on the bedding planes. When the beds are thus exposed, 
/the softer, mud-cracked layers commonly crumble away, leaving 
natural casts of the mud cracks as projecting ridges on the lower sur- 
faces of the sandstone layers. Obviously, the conditions that favor the 
formation and preservation of mud cracks are also ideal for the preserva- 
tion of footprints of animals that have crossed the soft mud soon after 
its deposition, and therefore mud cracks and footprints are frequently 



The most favorable places for the occurrence of mud cracks are on the 
flood plains of large rivers, the landward portions of great deltas, and the 
wide flat shores of shallow interior lakes that shrink or disappear during 
the dry season. In spite of its alternate wetting and drying, the littoral 
zone is unfavorable for the formation of mud cracks because the mud has 
not sufficient time to dry out thoroughly before the return of the tide. 
Mud cracks may, however, form to a limited extent at the upper mar- 
gins of estuaries where the spring tides reach only for a few days in 
each month. The mud cracks can form only when the sediment is 
exposed to the air and allowed to dry and shrink. They are generally 
lacking therefore in marine deposits. 

In the early Paleozoic formations, however, there are well-known 
examples of mud-cracked marine limestones. They represent very 

Fig. 156. Current-formed ripple marks exposed at low tide near Windsor, Nova 
Scotia. The movement of the current was from left to right. (Geol. Surv. of Canada.) 

special conditions that may find an analogy at present in the Runn of 
Cutch or in the Bahama Islands. During the hurricanes- of the spring 
of 1928 the lower islands and the borders of the larger islands of the 
Bahama group were flooded by the seas that swept over the Bahama 
Banks (see page 222). It was later observed that many square miles 
of this sediment were mud-cracked. In places limy sand and fragments 
of marine shells have since blown over it. It is quite possible that in 
the past similar conditions have permitted the covering of extensive 
low coastal lands by marine sediments laid down during exceptional 
hurricanes beyond the normal confines of the sea. 



Ripple Marks. Where currents sweep granular sediments along 
the bottom, the surface of the deposit develops parallel ridges resembling 
the ripples on the surface of a pool of water. These are known as 
ripple marks. They are also formed on the land where sand is shifted by 
the wind, as on sand dunes, or under running water. Ripples thus 
produced retain their form and migrate slowly with the current, because 
the sand grains are rolled up the windward or stoss side and fall down 
the leeward slope. Current-formed ripples, therefore, have an unsym- 
metrical form, the stoss side being a gentle slope and the opposite side 
steeper. The direction of flow of the current is thus autographed in 
the form of the ripple marks (Fig. 156). 

Where oscillatory waves touch bottom they also develop ripples as 
a result of the to-and-fro motion of the bottom particles; but oscillatory 
ripples, in contrast to current-formed ripples, are symmetrical (Fig. 157). 

Fig. 157. Unsymmetrical profile of current ripples (A) contrasted with the sym- 
metrical profile of oscillation ripples (jB) . 

Current ripples may be formed wherever currents disturb sandy or 
silty surfaces, whether it be wind currents on the land, the currents of 
running streams, or any of the currents in the seas. Waves of oscil- 
lation, on the other hand, occur only under standing water and in depths 
touched by wave action. Ordinary storm waves in the sea are ineffec- 
tive below 200 or 300 feet, but exceptionally the bottom is rippled to 
depths of 600 feet or more. 

Cross-bedding. In many deposits of coarse detritus, such as con- 
glomerate and sandstone, the layers of particular beds are inclined to 
the general planes of stratification at considerable angles (Fig. 158). 
This structure is known as cross-bedding. It is produced where sand is 
shifted by either wind or water currents in such a way as to be spilled 
down the front of an advancing deposit; as the foreset edge of a delta, 
the front of a gravel bar in a stream, the front of a sand dune, or merely 
the front of a current ripple. In any case the individual layers or lam- 
inae come to rest on a slope inclined to the general surface of deposition. 

The scale of the cross-bedding may be great or small. In dune sands 
single cross-bedded layers are commonly tens of feet thick but in silts 



shifted by small ripples the cross-bedded layers are fractions of an inch 
thick. Since the inclination or foresetting of the laminae is always 
downstream, it is possible to infer the direction of the currents that 
moved the cross-bedded deposits. 

The cross-bedding produced in dune sands is one of the most distinc- 
tive types (Fig. 158). The laminae are inclined first in one direction 
and then in another as a result of the repeated change of direction of the 
winds. Moreover, the front slope of the dune is not straight but curved 
where the sand rolls forward at its base as a result of eddies in the wind 

Fig. 158. Cross-bedding of the type characteristic of dune sands. Navajo sandstone 
naar Glen Canyon, Utah. (U. S. Geol. Surv.) 

on the leeward side of the dune. As a result, the cross laminae, devel- 
oped on a large scale, descend in curves that become tangential to the 
bedding planes below. 

Concretions. Stratified rocks in many places contain inclusions 
called concretions. These objects differ in composition from the en- 
closing rock and are generally rounded or nodular in form; some are 
quite spherical, others flattened, ovate, elongated, ring-shaped, or 
compound; still others exhibit odd and fantastic shapes. They range 
in size from a fraction of an inch to many feet in diameter. Ordinarily 
they are formed from one of the minor constituents of the rock; thus, 
in chalk and limestone they are composed of silica; in sandstone, of 
iron oxide or carbonate of lime; in shale, of calcium carbonate or sul- 
phide of iron. Although some are pure they commonly contain large 
amounts of the inclosing rock material, and the planes of stratification 
of the rock can be seen passing through some concretions. 



Their origin appears to lie in the solution of some of the minor con- 
stituents in the rock and the redeposition of this material around certain 
centers as nuclei. Very commonly they contain at their center a fossil 
and in such cases it appears that the products of decay of the buried 
organism have caused the precipitation of the mineral matter, the organ- 
ism serving as a nucleus. Remarkable imprints of fern leaves, insects, 
and marine animals are obtained by splitting such concretions. The 
shells and bones of even large animals are found in some concretions. 

Fig. 159. Concretions from clay beds, Long Island. 

Some concretions appear to have formed about nuclei of inorganic 
substance, as grains of sand, and still others show no definite nucleus 
of any sort. Iron oxide concretions are not uncommonly hollow or 
have only a partial filling of loose sand. 


Sedimentary rocks present a great variety of colors, some of which 
help to indicate the conditions under which the sediments accumulated. 
Sandstones are usually light-gray, greenish-gray, buff, brown or red; 
shales generally light-gray, dark-gray, greenish-blue, red, purple, ma- 
roon, or black; limestones are nearly white if pure but gray, buff, brown, 


pink to red, or black if impure. Some of the more recent deposits of 
clay are nearly white or have delicate shades of pink or lavender. The 
coloring matter is usually carbon or oxides of iron. 

Black or dark-gray sediments owe their color essentially to the black 
carbon resulting from the partial destruction of organic matter that was 
buried with the sediment. The black shales are especially rich in organic 
matter, and coal represents concentrated deposits of this nature. 

Oxides of iron are the great coloring agents in nature. Iron is the 
fourth most abundant element in the Earth's crust and iron-bearing 
minerals are disseminated throughout almost all rock masses. There 
are few sediments, therefore, that are completely wanting in irpn. 
In the igneous rocks the iron occurs generally in the ferrous state com- 
bined in the silicate minerals and in this condition is not strongly colored. 
In weathering, however, the iron becomes oxidized and then takes on 
bright hues of yellow, brown, 1 or red like those displayed by rusting 
iron. Iron oxide has the capacity of combining with a variable pro- 
portion of water to form hydrated iron oxides and the shade of color 
from yellow through brown to red seems to depend largely on the 
proportion of adsorbed water. The dehydrated ferric oxide (Fe 2 3 ) 
is of deep red color. 

Red sediments are commonly associated with beds of rock salt and 
gypsum that have been precipitated through the evaporation of large 
bodies of salt water. They also commonly make up or accompany 
sandstones and conglomerates containing unweathered feldspars. This 
common association of red sediments with the phenomena of arid regions 
has inclined many to regard all red formations as the products of arid 
climate. Against this belief there is, however, the anomalous fact that 
the great areas of modern red soil are in the warm and humid regions, 
and most of the deserts have dun or brownish soils. Since the sediments 
transported by streams are all recruited from regolith, it appears, there- 
fore, that the greatest sources for red sediments are warm and somewhat 
humid lands. However, red soils form where there is sufficient relief 
to insure free circulation of the oxygenated water through the regolith; 
and in the lower, swampy regions, vegetation is luxuriant, the stagnant 
waters are quickly robbed of their oxygen and here strongly reducing 
conditions exist. Therefore, even if red oxides be introduced into 
swampy, humid regions they tend to disappear and the sediments turn 
dark because of the excess carbon. Since sediments are generally 
deposited in the lowest places, it follows that in spite of their original 
color they have small chance of forming red deposits in humid basins. 
But life is sparse in the desert basins and the soil is dried and oxidized 
to great depths between the infrequent rains. No reducing conditions 


obtain here and if red sediments are washed in from the more humid 
uplands or slopes they are likely to remain red in color. 

Due to the abundance of marine life, reducing conditions generally 
obtain on the sea floors and, consequently, marine sediments are not 
commonly red. Nevertheless where red sediments are swept into the 
margin of the sea in sufficient quantity there may not be enough organic 
matter on the bottom fully to reduce the iron oxides. Off the mouth 
of the Amazon, for example, there is an area of modern red muds, and a 
number of cases could be cited among the sedimentary rocks where 
abundant marine fossils are entombed in red strata. The color of the 
abyssal red clays is apparently due to the sparseness of life on the ocean 
bottom and to the extreme slowness of accumulation which permits 
thorough oxidation of the fine sediments. 

In conclusion it should be evident that red strata occur chiefly in the 
continental formations and especially those that accumulated in arid 
or semiarid environments, but that red color is not of itself proof of any 
one condition of deposition. 


Much light is thrown on the geologic conditions under which a group 
of strata was formed by a study of the distribution of the beds, their 
relation to other rock masses, and, especially, the distribution of the 
fine and coarse sediments in the group. Some of the more important 
of these stratigraphic relations are discussed in the following paragraphs. 

Relative Age of Beds. In view of the origin of stratified rocks as 
superposed layers of sediment, it is evident that each layer is younger 
than the next below. Except where the rocks have been overturned, or 
broken and thrust out of their normal sequence as in some mountain 
zones, the youngest is at the top and the oldest at the bottom of any 
group of strata. Many conclusions as to the historical sequence of 
geological events rest upon this evident and fundamental law of strati- 

Grouping of Strata into Formations, A group of similar strata, 
closely related in their development, constitutes a geological formation. 
The view of the Grand Canyon wall seen in Fig. 160, for example, shows 
six great formations. The lowest includes intergrading beds of sandy 
shale and shaly sandstone more than 1000 feet thick; the next is a 
rather pure and homogeneous limestone that outcrops in 500-foot 
cliffs along the middle of the Canyon wall; the third is a thick group of 
shaly and sandy beds; the fourth a homogeneous shale; the fifth a 
pure, cliff-making sandstone; and the uppennost a limestone. During 



the time of the formation of the lowest group of strata, conditions of 
deposition must have remained nearly uniform and the processes supply- 
ing fine sand and mud to this region presumably operated with more 
or less continuity. A different set of conditions then came into existence 
and endured while the thick limy sediments of the second formation 
were produced. Eventually conditions changed again and yet again 

Fig. 160. The south wall of the Grand Canyon of the Colorado River near El Tovar, 
showing six great geological formations, each of which is marked with a letter in the 
explanatory diagram. (U. S. Geol. Surv., and Ariz. Bureau of Mines.) 

to give rise to the higher groups of strata. Thus each of the formations 
is the product of a distinct set of formative processes. 

Ideally a formation is a homogeneous group of strata, as limestone or 
sandstone or shale, but it may also be a group of interbedded layers of 
different kinds of rock, as the shale and sandstone in Fig. 148. All sedi- 
mentary rocks are divided by geologists into formations, but the forma- 
tional units are not generally so ideally simple and homogeneous as those 
shown in Fig. 160. For example, the 200 feet of strata exposed in the 
river bluffs about Kansas City are all so closely related that they are 
embraced in a single formation, although they include a series of alter- 



nating shale and limestone beds. Minor groups or subdivisions of 
such a formation are distinguished as members. Figure 161 shows five 
members of the Kansas City formation. Not uncommonly it is difficult 
to decide upon the limits of such formations and the grouping is more or 
less arbitrary. 

It is the universal custom to give rock formations geographical names. 
Thus, certain limestone beds well exposed about St. Louis are known as 
the St. Louis formation and the name is extended to them wherever they 
can be traced or identified. Likewise the beds forming the river bluffs 

Fig. 161. A portion of the Kansas City formation at Kansas City, showing five 
distinct members, two of which are of shale and three of which are of limestone. (Dunbar.) 

about Kansas City and extending far away north, east, and south, con- 
stitute the Kansas City formation. 

Intergradation of Different Kinds of Sedimentary Rocks. Sedi- 
ments vary from place to place in their coarseness and in their composi- 
tion. Gravel and coarse sand accumulate where transporting currents 
are able to carry them, but grade laterally into finer sand where the cur- 
rents slacken. Fine sand in turn grades into mud in the sheltered places 
and in quiet depths of water. The limy deposits of clear shallow sea 
bottoms grade eventually into muddy or sandy deposits near degrading 
lands. The change from one type of sediment to the next is always 
gradational and seldom abrupt. It is inevitable, therefore, that sedi- 
mentary rocks of one kind grade into contemporaneous strata of different 
composition. In general, coarse sediments grade most rapidly into other 


types of deposit and are most localized and irregular in their distribution. 
Thus conglomerates, which mark old stream channels or shore zones, 
generally form irregular linear deposits extended parallel to the stream 
course or the shoreline but in other directions grading into sandstone 
within a few miles at the most. Coarse sandstones generally show a 
similar inconstancy but single beds of shale or limestone may cover 
thousands of square miles. 

In a basin receiving sediments for a considerable time, gradations will 
occur not only in a lateral sense, but from bottom to top of the deposit, 

as the erosion cycle pro- 
gresses. If, for example, 
a land area were uplifted 
and then slowly degraded 
the sediments derived from 
it and deposited in adja- 
cent depressions would 

-Limestone 1 - 


... -. - --'-v Sandstone ryvT.. -- .... -. -.. ..- . , < , 

snow a regular cycle of de- 

position (Fig. 162), with 
Fig. 162. To indicate the normal succession of coarse material near the 

sedimentary deposits derived from a degrading land bage succeeded by g^ 

and finer sediment. The 

sedimentary cycle is a corollary of the erosion cycle. In the youthful 
and submature stages of dissection the streams can transport abundant 
coarse sediment but, as the land is worn lower, the streams flow more 
slowly and carry only fine material, and finally, late in the cycle of 
erosion when mechanical sediments are reduced to a minimm-r^ calca- 
reous deposits may form near the shore. If the land area is then reju- 
venated, rapid erosion begins again and a new cycle of deposition is 
inaugurated, clastic sediments coming to rest upon the limestone. 

Actually, the sedimentary cycle is seldom if ever brought to completion 
'in the simplicity pictured above, for it may be interrupted by many 
factors as, for example, uplift of the land, increased erosion due to cli- 
matic changes, uplift of the region of deposition or change in the course 
of marine currents. 

Overlap. Where the sea (or the area of deposition) gradually en- 
croaches upon a land surface, the beds have the relations shown in 
Fig. 163, each layer overlapping the next below and extending some dis- 
tance beyond it so as to thin out against the old land surface. This 
relation, known as overlap, is of common occurrence and great geologic 
importance. It shtfdTcPbe noted that the sands, being near-shore 
deposits, grade laterally into muds and that they are not of one age. 
At any place the sand is succeeded by mud but in successive sections 



from right to left the sand rises higher in the sequence as it follows the 
encroaching shore line. 

If, on the contrary, sea level should remain constant and accumulating 
sediments extend the land surface seaward, each bed may fall short of 

_ , Sea level C 

Fig. 163. To illustrate the progressive overlap of beds A, B, and as a result of the 
rise of sea level by three corresponding stages. Note that the shoreline has moved pro- 
gressively to the left. Symbols for sandstone and shale same as in Fig. 162. 

the next below as in Fig. 164, producing the relation known as offlag. 
It should be noted that in this case coarser sediment is carried out over 
finer as the sea floor is silted up. A gradual lowering of the sea level will 
also produce offlap as the accumulating sediments are then shifted 
farther seaward at successive stages of the retreat of the shore. 

Alternations of subsidence and building lead to an interfingering or 
dovetailing of finer and coarser sediments. The same phenomenon may 

Shore line F 

Shore line G 

Sea level E G 

Fig. 164 To show offlap of beds E, F, and G as a result of seaward building of the 
sediments while the sea level remained constant. Note that the shoreline has moved 
progressively to the right. A gradual lowering of sea level would produce similar results. 
Symbols for sandstone and shale same as in preceding figures. 

be produced by fluctuations in the rate of erosion or in the strength of 
currents. For example, during times of great floods or of stormy seas, 
coarse sediment is distributed farther than in times of quiet and may 
spread as layers of sand or silt over mud bottoms. Likewise the mud 
stirred up by exceptional storms may spread into regions of limy de- 
posits. The interbedding of thin layers of limestone and shale or of 
sandstone and shale, like that shown in Fig. 148, and in the shaly for- 
mations of Fig. 160, illustrates the results of such fluctuations. 


Extent and Form of Sedimentary Formations. In view of the dis- 
continuous nature of the regions of deposition it is evident that no 
geologic formation can be world-wide in its extent. Beds of mechanical 
sediment, as sandstone and shale, imply land surfaces from which they 
were derived, and basins in which they were laid down; obviously their 
areal extent must be limited by the borders of the basins next to which 
the sediments thin and disappear. In geometrical form a group of 
sediments is relatively broad and sheet-like, consisting of subparallel 
layers or strata. If deposited in a circumscribed basin the deposits will 
be roughly lenticular and thickest in the deepest (or most rapidly sub- 
siding) part of the basin; but if laid down along an open sea coast 
the form of the deposits tends to be wedge-like, thickest near shore and 
progressively thinner toward the ocean. 

Unconformity. One of the most important stratigraphic relations 
is that of unconformity, a subject that is treated in Chapter XII. 


Treatise on Sedimentation; by W. H. Twenhofel. 661 pages. Williams and 
Wilkins Co., Baltimore, 1926. 

A source book for the study of both sediments and sedimentary rocks, with many 
references to special literature. 


The various agencies that modify the surface of the Earth, such as 
the atmosphere, the water in its several forms as rivers, seas, and ice, 
and plant and animal life, derive the energy that enables them to do 
their work from a source exterior to the Earth: from the sun. For 
without the sun these activities would cease and the Earth's surface 
would be inert. Toward these agencies the Earth is passive, except 
as it adds the force of gravity to help them in their work. 

We have now to consider a set of agencies that also are modifying the 
Earth's surface, but whose energy is derived from sources within the 
Earth itself. So far as we can judge they are due either directly to the 
interior heat of the Earth or to changes going on within the Earth that 
produce heat. We shall describe first the results that they accomplish 
at the surface, and then inquire into their origin. 


General Description. A volcano is a hill or mountain composed of 
materials amassed around a vent through which they have been ejected 
from the Earth's interior in a highly heated or molten condition. The 
typical volcano is conceived of as a steep conical mountain with a pit- 
like crater at its top, from which issue from time to time gases, ashes, 
bombs, and flows of molten rock called lava. The ejection of material 
is termed an eruption, and to the human mind volcanic eruptions are 
perhaps the most impressive of all geological phenomena, from the im- 
mensity of the forces displayed, the magnitude of the results achieved, 
and the disastrous consequences that they frequently entail. Volcanoes 
vary widely from the typical form: some are low and of gentle slopes, or 
high and steep; conical, or elongated and irregular in shape; and the 
crater may be at the top, or on the side, and it may be of variable 
shape, or even be wanting. 

In size volcanoes range from small cones 100 or 200 feet high to some 
of the loftiest mountains on the globe. Thus certain of the highest 
peaks of the Andes are volcanoes; some of these are still active, as 
Cotopaxi in Ecuador, 19,600 feet high, with a crater half a mile in 
diameter and 1500 feet deep, whereas others, like Aconcagua (23,000 




.:>/:., i ijL^j^ 


feet) and Tupungato (21,500 feet) on the border between Chile and 
Argentina, and Chimborazo (20,500), in Ecuador, which apparently 
have no craters and are not now active, have become extinct in the recent 
geologic past. These volcanoes are built upon a dissected platform of 
much older rocks, above which they rise 10,000 to 12,000 feet; but the 
volcanoes of the Hawaiian Islands rise from the bottom of the Pacific 
Ocean at depths of 14,000 to 18,000 feet, and their highest summits 
project 14,000 feet above sea level, thus making the total height 30,000 
feet. In the United States the higher peaks of the Cascade Range, 
beginning on the north with Mt. Baker (10,703 feet), Mt. Rainier 
(14,408), and Mt. Adams (12,307) in Washington; Mt. Hood (11,225) 
Fig. 165, in Oregon; and Mt. Shasta (14,162) in northern California, are 
volcanoes, which are now dormant or have recently become extinct* 
Lassen Peak (10,460 feet), standing at the south end of the Cascade 
Range, is the only active volcano within the continental United States. 
Mt. Etna, on the coast of Sicily, rises 11,000 feet above sea level, and the 
diameter of the base of its cone is 30 miles. Its lower slopes are gentle 
and studded with many small, or " parasitic," cones. 

Character of Eruptions. Three kinds of material may be ejected 
from volcanic vents: gases, liquids consisting of molten rock, and solid 
material in the form of fragments; and the nature of a volcanic eruption 
depends largely on the proportions of these three things. If the eruption 
is violent and explosive, then the gases have been the chief factor in its 
production, and solid fragmental material is the main product; if, on 
the other hand, the eruption proceeds quietly, liquid rock, or lava, is 
the main product, and the gases play a less important role. We may 
roughly classify volcanic eruptions into those which are explosive and 
those which are quiet. When we classify actual volcanoes according to 
this difference in operation, we quickly find that, although there are good 
examples of both types, many, perhaps the majority, are intermediate in. 
their character; that is, at times they erupt violently and at other times 
they quietly discharge flows of lava. In many volcanoes during a 
quiescent stage there appears to be a gradual accumulation of pressure, 
the lava rises in the conduit, and eventually the eruption begins ex- 
plosively. Great quantities of gases mingled with dust and stones are 
ejected; and the pressure being thus largely relieved, the explosive phase 
is succeeded by a quieter one in which the lava escapes through rents 
in the cone and flows out on its exterior. 

Explosive Type. The most extreme volcanoes of this type give rise 
to appallingly disastrous explosions. Enormous quantities of gas are 
suddenly ejected into the atmosphere, so thickly mingled with commi- 
nuted rock (dust and ashes), as to form vast outrushiog and expanding 



Fig. 166. Eruption of Vesuvius, April, 1906. Seen from Boscotrecase. The vol- 
cano is 4000, the ash cloud more than 17,000 feet high. 



clouds of dense appearance and dark color (Fig. 166). The greatest 
known explosion was that of Tamboro, on Sanibawa Island east of Java, 
in 1815. Between 28 and 50 cubic miles of material were blown into 
the air. The ashes produced darkness for three consecutive days for 
a distance of 300 miles from the volcano and dust fell over an area of 
1,000,000 square miles. Krakatoa, a volcano in the Strait of Sunda 
near Java, exploded in August, 1883. After premonitory outmshes 
of gas for some time, the great explosions occurred, which blew into 
the air over a cubic mile of material in the form of dust and ashes. 
This rose as a vast dark cloud 17 miles into the atmosphere, by its dense- 
ness completely hiding the sun over an enormous area. The noise of the 
terrific detonations was heard as far as Australia; and the disturbance 
in the atmosphere was registered by barometers over the whole world. 

Fig. 167. Fiery cloud of Mont Pelee descending the mountain slope to the sea. The 
cloud at this moment is 7000 feet high and moving forward at the rate of over a mile in 
1.5 minutes. (A. Lacroix.) 

Huge waves, up to 100 feet above tide, were generated in the sea and 
rushed along the low-lying coasts of Java and Sumatra, sweeping far 
inland and destroying towns, villages, and the lives of nearly 40,000 
people; these waves were perceptible 3000 to 4000 miles away. 

In May, 1902, from the volcano of Mont Pelee on the island of Marti- 
nique in the West Indies, and almost simultaneously from that of Sou- 
fri&re on St. Vincent 90 miles away, after small premonitory symptoms, 
violent explosive eruptions took place. No lava was poured out, but 
the intensely heated gases were so thoroughly charged with incandescent 
particles of rock that the heavy, fiery clouds rushed down the mountain 
slopes to the sea. Destroying all life in its course, the cloud on Marti- 
nique swept through the town of St. Pierre and immediately destroyed it, 
together with its 28,000 inhabitants. On St. Yincent 2000, people 



perished and a broad tract of country was devastated. For many 
months after, Mont Pelee continued to eject at irregular intervals these 
incandescent clouds, one of which is seen in Fig. 167 rushing toward 
the sea. 

Intermediate Type. Probably most volcanoes belong, or have be- 
longed, to this class. Their eruptive periods are likely to begin with 
explosive activity, manifested by the ejection of gases in great quantity, 
accompanied by solid fragmental material bombs and ashes. In 
a succeeding phase liquid material issues; it may be ejected by the still 
issuing gases or it may break through the crater walls and produce out- 
flows of lava, sometimes of great volume. Finally the volcano becomes 
quiet, its energy for the time being exhausted; the lava column sinks 
down in the conduit and a period of quiescence intervenes before the 
next eruption. Although this sketch gives in a general way the suc- 
cession of events, it must not be supposed that all volcanoes of this class 
are alike in the character of their eruptions, or that the same one always 
passes through a similar set of phases at each eruption, for there is great 
variability in these respects. The main point is that volcanoes of the 
intermediate class at times are explosively active and at other times 
quietly emitiflows of liquid lava. 

Vesuvius, tbe longest and most studied and therefore the best known 
volcano in the world, belongs in this class. It occupies the site of an 
older volcano, which in the time of the Romans 
appeared to be extinct, for, although they rec- 
ognized its nature, they had no traditions of 
its having been active. In the year A.B. 79 
the volcano became active in eruptions that 
destroyed the towns of Herculaneum and Pom- 
peii on its seaward flanks. A great part of the 
older cone on the side toward the sea was 
blown away or engulfed, and in its place the 
new center of activity, the modern Vesuvius, 
began to be built. This building up has con- 
tinued until the present cone has become 4000 
feet high. Partly enclosing it lies the sickle- 
shaped ridge of Monte Somma, the remains of the rim cf the older crater 
(Fig. 168). Vesuvius is in a state of almost constant, relatively mild ac- 
tivity, with irregular periods of violent eruption. The last great erup- 
tion occurred in 1906 (Fig. 166). 

From the nature of the material composing their cones it is probable 
that the great volcanoes of the northwestern United States, previously 
mentioned, and now quiescent or extinct, belonged in the intermediate 

Fig. 168. Map of Vesu- 
vius and vicinity. 



class, as well as the active volcanoes of the Alaska Peninsula and their 
extension on the Aleutian Archipelago. 

Quiet Type. Volcanoes of this type give rise to quiet outflows of 
liquid lava, unaccompanied by the explosive disengagement of gases and 
the ejection of solid material as dust, ashes, and bombs. The lava of 
these eruptions is very hot and highly fluid. There is a more or less 
constant escape of gases from it, but without the catastrophic violence 
of the previous types. The best examples are in Hawaii. 

The island of Hawaii consists of a vast mass of lavas surmounted by 
five cones: Kohala; Mauna Kea, now extinct (13,800 feet high); 
Hualalai (8300), active in 1801; Mauna Loa (13,700), now active and 

Fig. 169. Halemaumau, north, pool. October, 1921, 


(Hawaiian Volcano 

some of whose lava flows have been 50 miles long; and Kilauea (4000 
feet). On the eastern slope of Mauna Loa, 20 miles from its summit, 
is the great crater pit of Eolauea, rudely oval in shape and 9 miles in 
circumference. In the floor of this pit is Halemaumau, a circular de- 
pression, which before 1924 was 300 feet in diameter, and was occupied by 
a lake of liquid lava, boiling and fountaining from the escape of gases 
(Fig. 169). The temperature of the lava ranged from 1000 to 1200 C., 
the higher temperatures prevailing at times of increased activity. 
In 1924 the bottom of the pit in Halemaumau dropped suddenly 700 
feet, and this subsidence was followed by explosive eruptions, the first 
since 1790. Avalanehing from the sides of Halemaumau began and 


enlarged the pit, so that it became 3000 by 3400 feet across and 1340 
feet deep. After the eruption the pit filled again with lava. 

Mauna Loa is the " monarch among modern volcanoes." It exceeds 
all others in its mighty size and in the magnitude of its eruptive activity. 
At times immense columns of molten lava play as fountains several 
hundred feet high and afford a spectacle truly sublime. Its crater is 
at an elevation nearly 10,000 feet higher than that of Kilauea. The 
outflows of lava are more likely to occur through its flanks than through 
the crater rim, and sometimes they take place below sea level. 

Relation between Volcanoes and Magmas. The molten rock-matter 
that originates within the Earth's interior and gives rise to volcanic action 
and volcanoes is known as magma. When this issues at the Earth's 
surface, the liquid material, and the rock produced by its cooling and 
solidification, is called lava. It must not be supposed, however, that 
the composition which a solidified lava shows, if determined chemically^ 
is exactly the same as that of the magma that yielded the lava. For 
the deep-seated magmas contain, in addition to the mineral substances 
of lavas, great quantities of gases, especially water vapor, which are dis- 
solved in them under pressure. As the magma rises to the Earth's 
surface and the pressure on it is consequently diminished, the gases es- 
cape, usually with more or less explosive energy, and give rise to the 
spectacukr features of volcanic activity. As the different types of 
volcanoes and of the lavas that they yield depend in large measure on 
the magmas producing them, it is necessary at this point to consider 
the nature and composition of these magmas. 

Composition of Magmas. As indicated above, the substances that 
compose a magma may be divided into two classes: 1, those which when 
hot are volatile and which mostly escape as gases and vapors, such as 
steam, carbon dioxide, hydrochloric acid, sulphurous vapors, during 
the congealing of the lava; 2, those which are non-volatile and remain to 
form the essential ingredients of the solid lavas. These fixed constitu- 
ents are silica (Si0 2 ) and the oxides of six metals, aluminum, iron, 
magnesium, calcium, sodium, and potassium. 

Silica is an important constituent of all magmas, and metallic oxides 
occur in all of them, but the particular metallic oxides present in differ- 
ent magmas range from almost nothing to considerable quantities. 
However, a kind of general rule governs the composition of magmas; 
without going into details, which are given in the chapter on igneous 
rocks, it may be said that the magmas, although forming a continuous 
chemical series, can be divided into two classes: one in which silica and 
the alkali metal oxides soda and potassa (NagO and K 2 0) pre- 
dominate, and the other hi which, conversely, lime (CaO), iron oxides 


(FeO and Fe 2 3 ), and magnesia (MgO) predominate. Magmas of the 
first class, on cooling slowly, crystallize into a mass of mineral grains 
that consist chiefly of alkalic feldspar, 1 with which quartz is commonly 
associated; the resultant rocks are called siliceous lavas. Rhyolite is 
the most abundant of the siliceous lavas. Magmas of the second class 
yield on cooling little or no alkalic feldspar, but abundant lime, iron, and 
magnesia minerals, such as pyroxene, calcium feldspar, and magnetite; 
rocks of this kind are termed basic lavas. As a rule the siliceous lavas 
are light-colored, whereas the basic lavas are very dark or black, and 
heavy because of their content of iron minerals. The siliceous lavas 
are sometimes termed acidic lavas because of the predominance of the 
acid-forming radicle (SiC^) in them; the basic lavas are so termed 
because of the predominance of the bases (lime, iron, and magnesia) 
in them. Basalt is by far the most common of the basic lavas. 

These characters and relations may be summarized in the following 

f a. Volatile substances: gases, e.g., water, C0 2 , etc. 

Magmas consist of \ b. Non-volatile substances: constituents that form the solid 
I materials of lavas. 


Chief constituents Resultant rock Chief minerals produced 

a. Much silica; abundant Siliceous lava Alkalic feldspar and quartz 

alkalies (light-colored) 

^ ^^ g -j- ca . ^^3^ Basic lava Pyroxene and calcium f eld- 

lime, iron, magnesia (dark-colored) spar 

Relation to Volcanic Eruption. The siliceous lavas, or rather the 
magmas that produce them, are, when their volatile constituents have 
escaped, thick viscous liquids, even at very high temperatures, as high 
as 2000 C. as experimentally ascertained. Parenthetically, it may be 
remarked that 2000 is far above the temperatures that prevail at 
existing volcanic vents, which are generally between 1000 and 1100 
and rarely reach 1200. The extreme viscosity is due chiefly to the high 
percentage of silica they contain, the amount in some kinds being as 
much as 75 per cent of the whole. For this reason when siliceous magma 
rises into the upper part of the conduit where the pressure is small and 
the contained gases begin to escape, the magma becomes stiffly viscous 
and the remaining gases can escape only with difficulty and usually with 
violence, giving rise to explosive eruptions. Hence volcanoes that yield 
siliceous lavas aTe likely to be of the explosive type, as Mont Pel6e. 
On the other hand the basaltic magmas, with about 50 per cent of 
silica, are very much more fluid, and they remain quite fluid down to 

1 For description of these and other minerals see Appendix A. 


much lower temperatures, probably down to 1000 C. or lower: the 
gases escape from them readily, but without explosive violence, as il- 
lustrated in the lava lake of Kilauea in Hawaii. Hence the basaltic 
magmas usually cause quietly eruptive volcanoes. 

The above statement indicates the general rule; it does not mean that 
basaltic volcanoes never have explosive eruptions, for a basaltic magma 
may become cooled in the conduit and in consequence be viscous, and 
thus permit its magmatic gases to escape only with difficulty and ex- 
plosive energy. The explanation applies chiefly to the two extremes 
and indicates what is probably the most effective cause for the explosive 
and quiet types of volcanoes. The intermediate type of volcano may be 
due in part to the intermediate kind of magma erupted from it, or to 
this factor combined with variations in viscosity at different "stages 
during the eruption as well as variations in the amount of gases. Sud- 
den accession of ground water into the volcanic pipe, as undoubtedly 
occurred at Kilauea in 1924, will produce an explosive eruption even 
in a basaltic volcano (see frontispiece). 


Gases. It has been shown already that the products yielded by 
volcanoes may be divided into three general classes: gases and vapors, 
solid fragmental material, and liquid lava. These products will be 
considered in more detail, beginning with the gaseous substances. 
The quantity of steam discharged by active volcanoes is immense, and 
is indicated by the height and volume of the cloud with which many 
eruptions begin. This cloud consists of the dust and ashes borne aloft 
by the uprushing column of gases. The great quantity of steam thus 
discharged into the atmosphere may give rise by condensation to 
heavy downpours of rain in the vicinity of the volcano; and, owing 
perhaps to the friction of the particles and to atmospheric disturbance, 
the eruptions and rains are accompanied by conspicuous electrical dis- 
plays and lightning. Although the composition of the gases during 
an actual volcanic eruption is not directly known, and it probably varies 
at different volcanoes and at different stages of an eruption, it is inferred 
with good reason from indirect evidence that it is chiefly gaseous water, 
or steam. As an instance of the quantity of water that some believe 
is discharged, Fouqu estimated that one of the subsidiary cones of 
Mount Etna discharged in 100 days in the form of steam the equivalent 
of over 460,000,000 gallons of water. 

In addition to the water, the different gases and volatile products 
exhaled by volcanoes make a long list. Not only are they given off from 


the vent itself, but the outflows of lavas continue for weeks and even 
months after their extrusion to emit gases as they cool and harden. 
Carbon dioxide, hydrochloric acid, hydrofluoric acid, and even hydrogen 
are given off, and to the mixture of the latter with oxygen and its sudden 
combustion are sometimes ascribed the explosions in the conduit. 
Sublimed sulphur and various compounds of sulphur, such as sulphur- 
etted hydrogen (H 2 S) and sulphur dioxide (S0 2 ), are emitted by some, 
but not all, volcanoes. 

Chlorides, especially ammonium chloride, are common at many vol- 
canoes in fact it was the abundance of chlorides at the Italian vol- 
canoes that suggested the idea that the eruptions are due to oceanic 
water leaking into the magma at depth but no chlorides occur at 
Kilauea, which is on an ocean-girt island. 

Fragmental Products. These are the volcanic projectiles, the 
materials blown into the air by the sudden liberation of the gases. 
They may be derived from the crust, or plug, of hardened lava left in 
the upper part of the conduit after a previ- 
ous eruption, from rock material torn from 
its walls, or from lava ejected from the 
upper part of the liquid column by the 
violent escape of the gases from the magma. 
Although the lava may start on its aerial 
flight in a liquid condition, it generally 
hardens in its passage and falls in solid 
form. The pieces of rock and the particles 
of magma driven upward and solidified are 
of all dimensions: from dust so fine that it 
may float in the atmosphere for several 
years, to large masses of several tons in 

weight. According to size, they are rough- *** 17 : " bomb ' 

ly classified as follows: pieces the size of 
an apple, or larger, are called blocks if ejected as solid fragments and bombs 
if ejected as particles of still-fluid magma; those the size of a nut are 
termed lapilli (meaning little stones); those the size of a pea are vol- 
canic ashes, while the finest is volcanic dust. The ashes and lapilli are 
frequently spoken of as volcanic cinders, and the cones made of them as 
cinder cones. It should be clearly remembered, however, that although 
these terms are used to describe the appearance of the products, the 
" cinders " are not products of combustion. A volcanic bomb is illus- 
trated in Fig. 170; it is of the variety called bread-crust bomb. 

The ejected products are in part composed of compact solid rock and 
in part are of a spongy, cellular, or vesicular character. This vesicular 


condition is due to the fact that, while the major part of the gases is 
passing into the air and carrying the fragments with it, a minor part 
is expanding in the particles of liquid, puffing them up into the cellular 
forms. Although the bombs, lapilli, and most of the ashes fall in the 
immediate vicinity of the vent and thus help to build up the cone, the 
dust may be carried by the prevailing winds long distances, hundreds 
of miles or more, and be thus spread over an immense area. Huge 
quantities are discharged in great eruptions, amounting to many mil- 
lions of tons (Fig. 166). Such dust showers may be very destructive to 
vegetation and even to animal life, but the soil ultimately yielded by 
them is very fertile. 

Liquid Material ; Lavas. In volcanoes whose periods of eruption 
begin explosively, the liquid lava generally issues later, after the vent 
has been cleared. The volcanic cone is not a structure of great strength, 
and is easily ruptured, or fissured, by the explosions and the pressure of 
the lava column, and hence the magma is not likely to flow out over the 
lip of the crater, but to issue through fissures in the sides of the cone. 
It may even happen, especially when the cone is composed of cinders, 
that, unable to withstand the pressure, one side gives way and allows a 
flood of lava to rush out through the breach thus made. 

The appearance and character of a lava stream and the material pro- 
duced by its solidifying depend on several things: on the chemical nature 
of the magma, on its viscosity, and the extent to which it retains its 
dissolved gases. On the chemical composition will depend the nature 
of the resultant rock, whether it will be a light-colored lava or a black 
basalt, or of intermediate character as previously explained. On the 
viscosity will depend the rate at which the lava will flow, the distance 
to which it will flow and in large measure the appearance its surface 
will present. When it issues, the lava is red or even white hot. It 
soon cools on the surface, darkens, and crusts over. If very viscous, the 
under part may yet be in motion, and the crust breaks up into a mass of 
rough, angular, jagged blocks of rock, which are borne as a tumbling, 
jostling mass on the surface of the slowly-moving flow. A typical 
Hawaiian flow of this kind advances with " a tremendous roaring, like 
ten thousand blast furnaces all at work at once." When eventually the 
flow comes to rest, the lava sheet is extremely rough and difficult to 
traverse. Such lava flows in Hawaii are called aa by the natives. 

In marked contrast other lavas may harden with smooth surfaces, 
which exhibit curious ropy, curved, wrinkled, or twisted and billowy 
forms, as seen in Fig. 171. Lava of this kind the Hawaiians term 
pahoehoe, in reference to its glistening, satiny surfaces. The difference 
between the two varieties of lava is determined in some way yet un- 



known by differences in the physical conditions during consolidation: 
for a flow may begin as pahoehoe and end as aa. 

Very fluid lavas move with considerable rapidity, as much as 10 or 
12 miles an hour, depending on the slope, Fig. 171; as they cool and 
become viscous the motion may be almost indefinitely slow, the stream 
creeping onward, possibly, for several years. 

Sometimes on slopes, after the lava has crusted over, the still liquid 
portion beneath may run out at the lower end, leaving long galleries, 

Fig. 171. Flow of basaltic lava running down a stream bed, the water of which is 
turned into steam. This lava, if cooling as seen, would have the pahoehoe surface. 
Hawaii. (XT. S. Geol. Surv.) 

tunnels, or caves beneath the crust. On some volcanic cones the natural 
drainage passes into these tunnels, disappears from view, and issues lower 
on the slopes in the form of springs. Such may be in part the cause of 
the springs around Mt. Shasta. 

Some magmas, or lavas, when ejected are too viscous to flow; they 
may then pile up over the vents in great domes. Such doming is chiefly, 
if not wholly, confined to the siliceous varieties of lava. Domes of lava 
have been observed in central France, Bohemia, Germany, etc., and are 
thought to have been formed in this way. They probably occur else- 
where. After the violent eruptions of Pelee, in 1902, the column of 



siliceous lava that filled the vent and had hardened into rock was pushed 
up so that it rose like a vast tower above the volcano, until it attained a 

maximum height of 1000 feet 
(Fig. 172). Gradually it crum- 
bled into a mass of blocks as a 
result of continuous explosions of 

Effect of Contained Gases; 
Vesicular Lava. That the lavas, 
even after they issue from the 
volcanic vent, still contain dis- 
solved gases is abundantly shown 
not only by the clouds of steam 
that may issue for weeks and 
months from them but also by the 
structures they assume as they 
cool into stone. Thus the upper 
part of a flow, especially of viscous 
lavas of the siliceous class, may be 
so p^ed up by the innumerable 

bubbles OI vapor in it expanding 

on relief of pressure that it may 

become a veritable glassy froth. Such rock froth, which is usually 
white or light-colored, is known as 
pumice , or pumice stone. 

In more fluid lavas, especially those 
of the basalt class, the bubbles are 
larger, and the resultant rock is 
spongy, cellular, or vesicular. This 
porous, eindery, or slag-Eke form of 
lava is called volcanic scoria. It is 
usually dark to black, or reddish 
(Fig. 173). Pumice, scoria, and other 
vesicular products are characteristic 
features of the upper surface of lava 
flows, and they constitute also a major 
part of the coarser fragmental mate- 
rials, such as bombs and lapilli, that 
build up the volcanic cone. 

Consolidation of Lavas; Glass and Stone. After lavas have been 
poured out and have solidified, most of them present the ordinary appear- 
ance of stone, but some, instead, have that of glass. The reason for this 

Fig. 172. -Rock tower of Mont 

Martinique. 1000 feet high. (A. Lacroix.) 

Fig. 173. Volcanic scoria. 


is as follows. If the liquid lava is not too viscous, the chemical molecules 
composing it will be capable of motion and will arrange themselves 
into definite compounds; that is, will crystallize into mineral grains, or 
crystals. Possibly the crystal grains thus formed will be large enough 
to be seen readily and the constituent minerals will then be determinable, 
or they may be so minute that the lava has a homogeneous appearance; 
nevertheless if crystallization has taken place the lava has the aspect 
of stone. On the other hand, if the lava is extremely viscous or quickly 
becomes so through rapid cooling, the molecules may not be able to 
arrange themselves, or crystallize, into minerals, and the mass solidifies 
as a glass. 

Thus while lavas in hardening into rock ordinarily take on a stony 
aspect, under certain conditions they may become glassy. Glasses 
form chiefly as the result of the freezing of siliceous lavas for, as previ- 
ously explained, they are usually the more viscous. Volcanic glass is 
called obsidian; certain varieties, in allusion to their luster and appear- 
ance, are called pitchstone. 

Some obsidians are pure glasses, others are mixtures of glass and crys- 
tals. In Yellowstone Park, Obsidian Cliff presents a section of volcanic 
glass 100 feet thick, which has cracked into columns in cooling. Such a 
thickness of purely glassy lava is unusual. It is chiefly on the edges and 
upper surface of lava streams that these glassy forms are, found. Primi- 
tive peoples, before they gained a knowledge of metals, made much use 
of obsidian for making knives, arrow and spear points, etc., in a manner 
similar to their use of flint. 


Kinds of Cones. The nature of a volcanic cone depends on the 
material of which it is built. If composed wholly of fragmental prod- 
ucts, the cone is high and steep in proportion to its size. This steepness 
is due to the high angle of repose for lapilli and volcanic ash; such 
material is very angular, rough, and clinging, and slopes of 40 are at- 
tained before the accumulating mass begins to slide. Cones of this 
kind are called cinder cones, and they are characteristic of volcanoes of 
the explosive class (Fig. 174). In contrast with them are the lava cones, 
formed entirely by quietly outflowing liquid lavas, like that of Mauna 
Loa (Fig. 175). They are necessarily very low and flat in proportion to 
their size, the angle of inclination being less than 10 (Fig. 175). These 
lava cones are built up by volcanoes of the quiet type. Most volcanoes, 
however, and this includes most of the largest volcanoes in the world, 
are of the intermediate type in their eruptive activity, and in conse- 



quence their cones have forms that are intermediate between those just 
described. For they are built up by the fall of ashes and lapilli when 
they are explosively active and by lava flows when they erupt quietly. 
The great cones of the Pacific States Mts. Shasta, Hood, Rainier, 
and others of the Cascade Range have been built in this way. 

The eruptions that break out on the lower flanks of the larger vol- 
canoes give rise to smaller, or " parasitic " cones. Mt. Etna is sur- 

Fig. 174. Cinder cone, showing steep angle of repose of lapilli. Outline as given by a 
photograph of Mayon volcano in the Philippines. 

rounded by over 200 of these, some of which are nearly 700 feet high. 
San Francisco Mountain in Arizona, an extinct and partly eroded vol- 
cano, has a number of such minor cones, some of them remarkably well 
preserved. As an active volcano grows, the earlier parasitic cones may 
be buried and concealed under later accumulations, or, in the declining 

Fig. 175. - 

- A lava cone, to show contrast with Fig. 17-4. From the Snake River 
plain, Idaho. 

stages of activity, the eruptive energy may show its last efforts in 
building them, as appears to have happened at the San Francisco volcano, 
just mentioned. 

Calderas; Explosion and Subsidence Basins. The term caldera, 
from the Spanish for caldron, is applied to crater-like basins of great 
size, especially those which are very broad as compared to their depth. 
The name is taken from the huge pit in the Canary Islands, called La 
Caldera, which is from 3 to 4 miles wide and bounded by lofty cliffs 
1500 to 2500 feet high, except on one side where the encircling wall is 
breached. From without, as seen from a distance, the general aspect 
of La Caldera is that of a huge cone truncated far below its apex. 

Many such great calderas occur in various parts of the world, and a 


study of them has led to the view that some of them were caused by 
gigantic explosions that have blown away a great part of the original 
cones as dust and ashes, leaving the calderas to mark their sites. Per- 
haps more generally they have been produced by the subsidence of the 
column of liquid lava in the conduit of the volcano, leaving a great 
cavity into which the superstructure of the cone has subsided. The 
truncated remnant of the cone makes the rim of the caldera. Prob- 
ably some calderas were formed by a combination of these two processes. 

Thus Tamboro in its explosive eruption of 1815, previously alluded to, 
blew away a good part of the original volcano and produced a caldera 
nearly 4 miles in diameter. 

The finest caldera in the United States is at Crater Lake in southern 
Oregon. This marvelously beautiful lake occupies a caldera at the 

Fig. 176. Part of the basin and wall of Crater Lake, Oregon. Note the small cone 
within (Wizard Island). (U. S. GeoL Surv.) 

summit of a volcanic mountain in the Cascade Range, and is 6 miles 
long by 4 broad, 2000 feet deep, and encircled by steep cliffs 500 to 2000 
feet high (Fig. 176). An island in it, made by a small but perfect cone 
of volcanic material, indicates a feeble renewal of eruptive activity after 
the principal subsidence. The caldera, if emptied of its water, would 
appear as a great basin. The reason for believing that the caldera was 
formed by the collapse and engulfment of the greater part of a former 
cone as the result of the subsidence of the lava column in the pipe of the 
volcano rather than by explosion lies in the absence of the debris 
about 18 cubic miles of material that so gigantic an explosion would 
have spread over the adjacent outer slopes. The former mountain, to 
which the name Mt. Mazama has been given, is conceived to have had 
the size and general character of Mt. Shasta and to have been heavily 
capped with snow and glaciers during the Glacial Period. 


Many craters and calderas of extinct or dormant volcanoes are filled 
with water, giving rise to lakes. Several of the circular lakes of Italy 
surrounded by ejected volcanic material, like Bolsena and Bracciano, 
are regarded by some geologists as marking the sites of great calderas. 

Explosion Pits. In some places where volcanic activity has begun 
it has proceeded no further than the initial explosions that drilled a vent 
through the country rock. The material blown out has made a low slight 
ridge around the pit, but no real cone was built. Sometimes volcanic 
products, such as pumice and cinders, are mixed with the fragments of 
the country rocks. Such basins range from a few hundred feet to several 
miles in width, and in humid regions they are usually filled with water 
and form lakes. Some of the best examples of them are in the region 
west of the Rhine in Germany, known as the volcanic Eifel. They are 
called maars (German, maaren), like the Pulvermaar. A pit that 
strongly resembles a maar exists at Coon Butte in Arizona. The basin 
sunk in the plain is f of a mile in diameter and 500 feet deep. The 
presence of meteoritic iron in and about it and other features have led 
to the view that it was caused by the impact of a huge meteorite and is 
probably not of volcanic origin. Hence it has recently been renamed 
Meteor Butte. 

Rebuilt Volcanoes. Not infrequently after a caldera has been 
formed, either by subsidence or by explosion, or both, a revival of vol- 
canic activity starts building up a new cone within it. This is shown on 
a small scale at Crater Lake, where Wizard Island represents the un- 
submerged top of a new volcanic cone that stands on the floor of the 
caldera formed by the collapse of Mt. Mazama. One of the best ex- 
amples is at Vesuvius, which has built itself up within the caldera of 
Monte Somma, as explained previously. From this rebuilding within 
an old crater or caldera there results a cone-in-crater structure, of which 
there are many examples. The vast crater-like pits that are so com- 
mon on the surface of the moon frequently show this arrangement, 
suggesting an analogous origin for them. Conceivably Vesuvius, be- 
fore it becomes extinct, may go on increasing in size until the old caldera 
is obliterated; it would then be a completely rebuilt volcano. 


Structure of a Composite Cone. If a column of magma is forced 
upward through the crust of the Earth until it reaches the surface, the 
relief of pressure will enable it to commence discharging its dissolved 
gases and vapors. Conceivably the pressure of the contained gases may 
be too great for the topmost layers of the bedrock to restrain them until 



the magma reaches the surface; consequently these layers may be blown 
into the air and a vent drilled ahead of the rising column of lava. Ar- 
rived at the surface, the magma may flow out quietly, or, if it is too 
viscous to do this, explosions may continue and material be blown 
upward. By the falling of the fragments around the vent a cone is 
built up, somewhat as seen in the diagram, Fig. 177. The pieces cannot, 
of course, fall back against the uprushing column of gases and cover the 
vent; they must fall outside of 
it, the heaviest and largest first 
and nearest to it, the smaller and 
lighter later and farther away, 
the distribution of the lighter 
depending much on the wind. 
Thus the cone grows as a circu- 
lar ridge upon whose crest most 
of the ejected material is deposi- 
ted. This material tends tO roll Fig. 177. Ideal section through a volcano. 

and slide both outwardly away J he dar1 ^ ! ay ^ ** the ^ n , es are /* buried 

J J flows or injected masses (dikes and sills). 

from the center and inwardly 

down the crater toward the vent. This process forms the cone and 
crater, and certain features of their structure follow as a consequence 
of this mode of formation. 

Tuff and Breccia. The deposits of successive eruptions are marked 
by layers, some of coarser, some of finer material, in each of which, if 
not composed of uniform-sized particles, there is a gradation from coarser 
at the bottom to finer at the top. Thus there arises a rude stratification, 
or bedding, the beds sloping down and out from the crater edge (Fig. 
178). The bombs, lapilli, and ash composing the beds gradually become 
compacted by their weight and by the infiltration and deposition of 
cementing substances. They are thus transformed into a more or less 
friable, porous rock called, when composed of the coarser materials, 
volcanic brecda, and when of the finer dust and ashes, volcanic tuff. In 
the crater the fragments are larger, generally large blocks of rock, and 
they usually form a tumultuous mass without order or arrangement, 
with intermingled finer material; such material is called volcanic 

Lava Flows and Dikes. In addition to the beds of tuff and breccia, 
liquid lava flows down the outer slopes of many volcanic cones, and 
when these streams harden into solid rock they protect the softer layers 
of tuff and breccia from erosion and give strength to the edifices. Since 
the lava rarely flows over the lip of the crater, but, especially in high 
cones, breaks through fissures on the sides of the cone, these fissures also 



become filled with lava, which hardens into rock. These rock-filled 
fissures are called dikes, and like ribs they also serve to strengthen the 
volcanic cone. Thus a vertical section through a volcano (Fig. 177) 
shows a central core of igneous rock surrounded by beds of tuff and 
breccia with intercalated flows of lava, which are cut by a radial system 
of dikes. This description gives us an idea of the general structure of 
a typical composite cone, one formed by the intermediate type of vol- 

Fig. 178, Inclined beds of volcanic ash. Part of a former cone at Trinchera, Colo. 

(U. S. Geol. Surv.) 

cano ; there are of course many variations from this, as may be inferred 
from what has been previously stated. 

Dissection of Volcanoes. At every stage of its existence a volcano 
is subject to the agencies of erosion and weathering, which tend to cut 
down all prominences on the Earth's surface. Its height and appearance 
at any given time are the result of the balance between these destructive 
forces and the upbuilding power of volcanism. Even active and grow- 
ing volcanoes are commonly trenched and scored by ravines and gulches. 
After eruptions, when the cones are covered with fresh deposits of dust 
and ashes, the latter become so saturated with water from the rainfall 
that they slide down as flows of liquid mud, leaving ravines, which are 
enlarged by subsequent storms (Fig, 179). 



As soon as a volcano becomes extinct, the ravages of erosion are un- 
checked, and the period of dissection ensues. The lighter tuffs and 
breccias are carried away more easily and rapidly; the harder, more 
compact, and resistant flows and dikes of lava, and the parts protected 
by them, are eroded more difficultly and slowly. It is surprising, how- 
ever, for how long a time some cones that are composed of mere cinders 

Fig. 179. Vesuvius in 1906, showing trenching by ravines in the ashes after the great 
eruption. (F. A. Perret. Courtesy of Harper's Weekly.) 

loosely piled will resist erosion and retain their form. The reason for 
this resistance to erosion is due to the porosity of the material, which 
allows the rainfall to sink through it without causing downwash. 

As erosion progresses, the mass of rock formed by the solidification 
of the magma in the central conduit is brought to view, provided the 
magma column was not drained off before the volcano became extinct. 
If the central column of magma was drained off, the site of the vent 
probably is marked by a mass of agglomerate. As erosion progresses 
and the cone is demolished, the cen- 
tral rock mass, owing to its greater 
resistance, is likely to form a decided 
prominence, and, even when erosion 
has finally swept away aU external ^ 1a _ Section through a partly 

evidence Of the COne and bitten ero ded volcano, with volcanic neck (6) left 

deeply into the underlying rocks, it 

may remain projecting, a monument 

to the vanished volcano. A rock mass of this kind filling a former con- 

duit is termed a volcanic neck (Figs. 180 and 193). Here we must pause, 

for this carries the dissection of the volcano to its very root. 

Extinct volcanoes occur in every quarter of the globe and in many 
regions where volcanic activity has long since disappeared. Every stage 
of dissection is represented among them: from cones only slightly worn 


to those so thoroughly eroded that the original shape has been entirely 
lost, but whose central rock core, outlying concentric masses of lavas, 
tuffs and breccias, and radial dikes still plainly show their former exist- 
ence. The Rocky Mountains, once a theatre of active volcanism, in 
many places are strewn with the wrecks of former volcanoes. Many 
occur in the Yellowstone National Park and surrounding country, where, 
as we shall see later, the spark of volcanism still lingers. So deeply 
eroded are the volcanoes that their remnants now form a region of most 
varied and irregular topography. 

Volcanoes and Deep Masses of Magma. In the preceding para- 
graph there has been sketched the structure of the volcano down to a 
conduit filled with magma derived from below. We cannot leave the 
subject of volcanic activity and the structures it gives rise to before 
stating that volcanoes are only one phase, the surface manifestation, of 
the general movement of masses of magma from unknown depths into 
the upper region of the Earth's crust. Although these masses of magma 
may attain the surface and there produce volcanoes or lava flows, many 
of them have not reached the surface, but have remained in depth in- 
truded into the rock layers of the cnist, and there cooling and solidifying, 
gave rise to rock bodies of varied shapes and of sizes from a few feet in 
thickness up to miles in extent. Such deep bodies of magma form what 
are known as intrusive masses of igneous rock, and every extinct volcanic 
conduit, if it could be traced downward, would be found to join some such 
intrusive mass below, or, if active, to extend into a body of magma which 
in time will solidify as an intrusive mass. A proper understanding of 
the intrusive masses demands a more extended description and explana- 
tion, which will be found in the next chapter. 


Age of Volcanoes. The span of life of active volcanoes differs greatly 
in different individuals. We know from written testimony that Etna 
has had the same general eruptive character for the last 2500 years. 
We estimate that because of its great volume it has taken at least 
300,000 years to build this grand volcano. It is certain that the erup- 
tion of such vast masses of material as make up the larger volcanoes 
must have required, from the human standpoint, an immense lapse of 
time. On the other hand, we know from various considerations that the 
present active volcanoes are from the geological standpoint recent 

It is also difficult to say whether a volcano is extinct, because long 
periods, hundreds of years, may elapse between eruptions. In the 


Middle Ages, Vesuvius had been so long dormant that its crater was 
overgrown with vegetation and it gave no sign of life. But in 1631 
it became violently eruptive, and has since been intermittently active. 

New Volcanoes. Within the period of recorded human knowledge 
a number of volcanoes have begun their existence, and many of them 
are still active. Vesuvius is, of course, the most noted case of this, but 
other examples are Jorullo hi Mexico, which came into being Sept. 
28, 1759, in the midst of a cultivated plain, and is now about 4300 feet 
high, and Izalco in Salvador, which began in 1770, has been almost 
continuously active since then, and is now over 6000 feet high. 

No well authenticated volcanic eruption was witnessed within the 
limits of the United States proper until May, 1914, when explosive 
eruptions, which have since continued at intervals, began at Lassen 
Peak in northern California. The eruptions have been chiefly of gases, 
ashes, and stones. In 1915 there were two enormous explosions 
two horizontal blasts that laid waste a wide swath of country. Lassen 
Peak, as already mentioned, is the only active volcano within the 
continental United States. 

Distribution of Volcanoes. The present active vents are about 430 
in number, and those cones which because of their slightly eroded con- 
dition may be considered dormant or only recently extinct amount to 
several thousand. They have a general tendency to be grouped in long 
belts on the Earth's surface. The most marked of these belts is the 
great zone that borders the Pacific Ocean; it passes northward along 
the Andes, through Central America into Mexico, through the United 
States and Canada to Alaska, then along the Aleutian chain to Asia, 
and turning southward through Kamchatka, Japan, and the Philippines, 
it crosses the East Indies, and by various island chains again passes into 
the Pacific. Certain portions of this belt, like the Andes and the 
Aleutian chain, are remarkably linear and well developed. Another 
great belt has an east and west direction: from Central America it 
extends through the West Indies; it then continues through the Atlantic 
by the Azores, Cape Verde, and Canary islands, runs through the Medi- 
terranean, through Asia Minor and Arabia, and continues along the chain 
of the East Indies, where it crosses the circum-Pacific belt, and extends 
out into the Pacific. This linear arrangement occurs not only on a large 
scale, affecting series of volcanic groups, but it occurs on a small scale 
as well, influencing the distribution of the volcanoes that compose the 
individual groups (Figs. 181 and 182) 

Volcanoes occur both on the continents and in the oceans, the true 
oceanic islands appearing to be entirely volcanic. Notably in the 
Pacific there are great numbers of them, many extinct or dormant, some 



still active, and here again many of the volcanoes are grouped in lines 
and stand on the submarine ridges which rise from the ocean floor. 
From the fact of linear arrangement has been drawn the important 
deduction that volcanoes are in general situated on, or near, lines of 
fracture, folding, and weakness in the Earth's crust. 

Lines of fracture and weakness have undoubtedly proved favorable 
sites for volcanic action, not only for a time, but in places for long- 

Fig. 181. Map showing distribution of actrve or recently extinct volcanoes in the 
Eastern Hemisphere. On S. L. PenfielcTs stereographic projection. 

continuing geologic periods, and thus they have greatly influenced the 
origin, situation, and arrangement of volcanoes. But, on" the other 
hand, it seems clear that a volcano, or a group of them, may originate 
where no definite connection between them and any fracture line can be 
shown to exist. And in places no tendency to a linear arrangement in 
the group may be seen. The volcanic forces appear to have been suffi- 
ciently powerful to find an outlet without needing the aid of a fracture. 



A good example of this may be seen in the Highwood Mountains, a 
group of extinct and greatly eroded volcanoes situated on the great 
plain of central Montana. While the remaining tuffs, breccias, lava 
flows, and dikes composing this group and their arrangement and 
attitudes indicate clearly the cones that once existed, erosion has dis- 
sected them so deeply that the shapes of the cones have been destroyed, 
the central conduits now filled with the massive rock are exposed, and 

Fig. 182. Map showing distribution of active or recently extinct volcanoes in the 
Western Hemisphere. On S. L. Penfield's stereographic projection. 

their relations to the sedimentary bedded rocks through which they were 
forced laid bare. The crust shows no evidence of profound breakage 
or displacement that might have determined the positions of the vents, 
nor do the conduits of the different volcanoes show linear arrangement. 
A striking instance of how little influence surface topography may 
have in determining the site of volcanic action, which in the immensity 
of its power appears to disregard such minor considerations entirely, 


can be seen at the Grand Canyon of the Colorado. Uninfluenced by 
its 5000 feet or more of depth, volcanoes have broken out upon its very 
rim, instead of in its depths, and their lavas have flowed down into it. 

That almost all active volcanoes are either situated in the sea, or in 
a general way around its borders, and when inland are in or near lakes, 
has led many to believe that there must be a necessary connection be- 
tween the surface waters and the cause of volcanic activity. This 
question will be considered later in discussing the origin of volcanoes/ 

Submarine Eruptions. From the great number of volcanic islands 
in the sea, the real oceanic islands being of this nature, it is evident that 
in times past tremendous eruptions and vast outpourings of lava have 
occurred on the sea floor. The volcanic chain of the Hawaiian Islands 
is an example of this. Actual eruptions beneath the sea have been ob- 
served and recognized by the issuance of vapors and ashes from the 
water. Thus, in 1831 a volcano was thrown up in the midst of the 
Mediterranean Sea, forming a new island called Graham's Island. 
Being composed of light cinders, it was soon destroyed by the waves 
and reduced to a shoal. The three Bogoslov volcanoes of the Aleutian 
chain formed in 1796, 1883, and 1906 are other examples. 

Such eruptions have occurred repeatedly in the past and their prod- 
ucts, mingled with sediments from the land, have been laid down as 
deposits on the ocean bottom, as seen in many places where the sea 
floor with these deposits has since been raised and become a land surface. 
Nor do these volcanic products differ essentially from those which are 
formed by volcanoes on the land. It also is probable that many of the 
cones formed beneath the sea, and thus protected from erosion, are of 
great age, even quite old from the geological standpoint, and have 
served as the foundations for coral islands, as previously discussed. 

Fissure Eruptions. Outflows of basaltic lava have taken place in 
several regions on such a gigantic scale and deluged such immense tracts 
of country that they cannot be referred to the outpourings of any vol- 
canic cone or group of volcanoes. Moreover, the cones from which they 
might have come are apparently wanting. These great lava floods have 
issued from fissures in the Earth's crust. The result of such flooding 
is that broad plains, or plateaus, consisting of successive level sheets of 
basalt lava, in places interlaid with beds of tuff, have been formed. 

Basalt plateaus of this origin are the great lava fields of the Columbia 
and Snake rivers in the far Northwest of the United States, which cover 
from 200,000 to 250,000 square miles and are in places 4000 feet thick; 
the Deccan traps (basalts) of western India, which are at least 200,000 
square miles in extent and reach a maximum thickness of 6000 feet; 
the northern British Isles, which in part, with the outlying island groups, 


appear to have been carved by the sea from a great basalt plateau that 
may have extended to Iceland. The horizontal layering of these lava 
sheets is evidently due to the extreme fluidity of the issuing magma, 
which permitted it to flow for many miles before congealing. Thus in 
Iceland, a lava flow has been traced a distance of 60 miles. It is such 
outpourings, which occur in other regions as well as in those mentioned, 
that exhibit to us the grandest effects of volcanism. It is conservative 
to say that since a comparatively recent geologic period as much as 
500,000 cubic miles of molten material have been transferred from the 
inside to the outside of the globe by the extrusive process of volcanism, 
most of it by fissure eruptions. 

Areal Eruptions. The theoretical possibility has been .pointed out 
that eruptions might occur through great openings formed in the crust 
by the foundering of the roofs of large magmatic reservoirs. Lava 
might thus well out on a stupendous scale. Some examples of this mode 
of eruption are thought to occur but have not been definitely established. 


General Remarks. So far as regards the nature of volcanoes, the 
character of their eruptions and of the products afforded by them, their 
distribution, and in some measure their life, we are dealing with ascer- 
tained facts. We know also quite clearly the reason for the different 
kinds of eruption and the three types of cones. But when we seek to 
learn the cause and origin of volcanism we must then consider the depths 
of the Earth itself, about which we know very little. We are led from 
facts into almost pure speculation, and this should be clearly understood 
by the student. It is evident that our ideas of the cause of volcanic 
action will depend on those which we have concerning the nature of the 
Earth's interior; what has been learned regarding the Earth's interior 
will be considered in a later place. There are, however, certain phases 
of the problem of volcanism that may be considered here. 

Problems of Volcanism. Some important questions that arise 
when we seek to discover the cause of volcanism may be stated as fol- 
lows: What is the origin of the heat that keeps the magmas in a molten 
condition? What is the origin and history of the magmas that come to 
the surface? From how deep down do these magmas come, and where 
is the seat of volcanism? What is the origin of the gases; have they 
always been contained in the magma, or have they been absorbed from 
outside sources, and, if so, when and where? And finally, what causes 
the magma to ascend to the surface from the depths and thus give rise to 
volcanoes? These are fundamental problems, most of which our knowl- 


edge at the present time is too limited to enable us to solve. Neverthe- 
less, the views held regarding them may be briefly stated and discussed. 

Origin of the Heat. At the present time the most prevalent view 
regarding the source of the heat necessary to produce the magmas is 
that it is original, residual from a globe once molten and still intensely 
hot in its interior. Some, however, regard the heat as due to the gradual 
contraction and compression of the Earth through the force of gravity. 

Some have urged that the crushing together of the Earth's outer shell 
through contraction must generate heat on an enormous scale. Such 
compression and crushing have taken place during the formation of 
mountain structures, as we shall see later, and it is inferred that through 
this process melting has occurred and volcanoes have been made. 
There are two objections to this view. The first is that many volcanoes 
occur where there has been no crushing of the outer crust, or at least, 
not for an immense period of geologic time antedating the appearance 
of the volcanoes, as at San Francisco Mountain and other volcanoes 
on the high plateau in Arizona. The other is that the folding and com- 
pression of the Earth's crust that makes mountain ranges is probably a 
very slow process, and although great quantities of heat would undoubt- 
edly be generated, it has not been shown why it would not be as rapidly 
dissipated or absorbed in doing chemical work. How could it become 
accumulated and concentrated sufficiently to produce melting and vol- 
canoes? For, to use an illustration, what we need is not a cask of warm 
water, but a cupful of boiling water. 

Inasmuch as the discovery of radioactivity has shown that some ele- 
ments are spontaneously disintegrating and breaking down into other 
substances, as for example, uranium into helium and lead, with the con- 
current production of heat in notable quantity, it has been suggested 
that the changes of this nature that are going on within the Earth 
produce the heat necessary for volcanic action and even for such vast- 
scale igneous manifestations as are represented by the enormous in- 
trusive bodies beneath the surface. 

Origin of the Magma. This is a complex problem. Different vol- 
canoes erupt lavas of different kinds. Vesuvius erupts one kind, Etna 
another. Furthermore, many a long-lived volcano has erupted a variety 
of lavas. For example, Lassen Peak, which is regarded as an old and 
dying volcano, has erupted rhyolite, dacite, andesite, and basalt. Did 
the supply pipe of Lassen Peak tap four different reservoirs containing 
four different magmas? It is believed that originally there was a 
homogeneous magma beneath Lassen Peak and that by a series of 
internal changes this magma yielded the other magmas. This proc- 
ess by which a magma of initially homogeneous composition splits 


up into unlike fractions is called magmatic differentiation. It is 
to this process that appeal is made to account for the diversity of 
igneous rocks, not only at Lassen Peak but at all other volcanic cen- 
ters. The causes that produce magmatic differentiation are many, 
but need not be explained here. It is thought that the initial primi- 
tive magma the world over is basaltic in composition. One of the 
cogent reasons for this belief is that the great fissure eruptions that 
have occurred in such enormous volumes at intervals throughout the 
whole span of geologic time are of basaltic composition; and this is held 
to indicate that a basaltic substratum of potentially liquid magma 
underlies everywhere the visible crust. The prevalent view is that the 
magma is a remnant of the original molten substance of the globe. 
Those who hold this view do not claim, however, that it has necessarily 
always been in a liquid condition. In melting, rock material expands; 
if sufficient pressure be put upon it, it cannot expand and, therefore, 
cannot melt. It is assumed that because of the tremendous pressure 
reigning in the Earth's depths the material although very hot is solid, 
but should the pressure be relieved at any place, as for instance, by 
upward buckling of the Earth's crust or by reduction of the superin- 
cumbent load as the result of deep erosion, or by both, then melting 
would ensue and a body of magma would be formed 

Origin of the Gases. The chief magmatic gas is water, as a rule 
making up more than 80 per cent of the total. This water may have 
been part of the original substance of the Earth, in short, of primitive 
origin " telluric water " (from Tellus, the Earth) or formed from 
the combustion of primitive hydrogen, or it may have been atmospheric 
water absorbed by the magma from the surrounding rocks, or it may 
have been acquired by the magma by melting up rocks containing water- 
bearing minerals. A quantitative evaluation of these possibilities is 
beyond the present powers of science. A magma that has dissolved 
much limestone would become highly charged with carbon dioxide, which 
is one of the important volcanic gases. Such absorption of limestone 
would generate enormous pressures, and such local development of 
pressure, it has been ingeniously suggested, may have determined the 
sites of certain volcanic vents. 

The fact that most volcanoes are situated in, or near, the sea or lakes 
has been considered a strong proof that the gaseous water contained in 
the magma has been obtained from descending surface waters. But 
this argument when examined loses its force. The nearness of some 
volcanic chains to the sea, like those of North and South America, is 
only relative to the size of the continental masses. Actually they are 
long distances inland: in South America from 100 to 250 miles and this 


includes some cones still active like Cotopaxi which are not near 
any inland water body; in North America from 30 to 100 miles or more, 
and, although these are mostly extinct, it can yet be shown that when 
active there were no bodies of water near most of them. 

Cause of Ascension. The fact that the great volcanic chains are 
situated on those belts along which movement and disturbance of the 
crust have taken place is significant. For these belts are apparently 
zones of weakness in the crust, and have thus afforded favorable places 
for the upward movement of the magma and its escape to the surface. 
As will be explained more clearly later, the Earth's crust is divided into 
great blocks, or segments, and these have in times past moved up or 
down with respect to one another. It is noticeable in many places that 
where one of these great blocks, measuring hundreds or thousands of 
square miles in area, has sunk, this subsidence has been attended by 
uprise of magma, outflows of lava, and commonly by volcanic action. 
Examples of such crustal subsidence and concomitant volcanic activity 
are the depressed tracts that form the great Rift Valley of East Africa 
and the valley of the Rhine. The mechanism of ascension of the magma 
in this way is likened to the action of a hydraulic press. 

The old idea that the Earth has a hot, liquid interior and that the 
downward pressure of the cold and solid crust collapsing on the shrink- 
ing nucleus forces this liquid out and thus gives rise to volcanoes has 
been completely disproved by a number of considerations and is no 
longer held. The independent eruptions of adjacent volcanoes in the 
same group, and the fact that the lava column in Mauna Loa stands 
10,000 feet higher than that in Kilauea, only 20 miles away, are disproofs 
of this view, and others will be mentioned later. 

As to the seat of volcanic action, or the point from which the magma 
may be considered to move on its upward way, this appears to differ 
in volcanoes from that of fissure eruptions. Seismic evidence, from the 
shocks attending volcanic eruptions, indicates that the magmatic hearth 
or reservoir tapped by the volcano is at a relatively shallow depth 
a few kilometers. Fissure eruptions, however, appear to tap the basaltic 
substratum, which on seismic and other evidence is believed to lie at a 
depth of 30 kilometers or more. 


Introductory. In the foregoing description of volcanoes it has been 
shown what an active role gases, especially superheated steam, play 
in their eruptions. But long after a volcano has ceased to be active 
and has passed into a dormant or dying stage these gases continue to 


issue from its crater, or from its flanks, or even from places in the sur- 
rounding country. In the same way thick beds of extruded lavas con- 
tinue, often for years, to exhale steam and other vapors. And, as we 
shall show later, it has often happened that bodies of magma have 
penetrated into the outer shell of the Earth without attaining the sur- 
face or forming volcanoes, and in solidifying they, like the lavas, have 
given off quantities of gases, which work their way through fissures up- 
ward to the Earth's surface. It is now proposed to describe the phe- 
nomena produced at the surface by such emanations. They may ap- 
pear as vapors or in liquid condition; the gaseous emanations may be 
considered under the general heading of fumaroles, the liquid emanations 
under hot springs. 

Fumaroles. This word, which is derived from a Latin verb meaning 
to smoke, is applied to fissures or holes in the rocks from which steam 
and other gases escape with more or less force. The " smoke " of the 
f umarole is thus mainly steam. Although steam predominates, generally 
forming 99 per cent of the total, other gases, such as carbon dioxide, 
hydrochloric acid, hydrogen sulphide, hydrogen, methane, and others 
also occur. Fumaroles that give off sulphurous vapors are termed 
solfataras, from the Italian word for sulphur. 

In addition to the substances already mentioned, fumaroles carry 
certain metallic constituents such as iron, copper, and lead. These 
metals have been rendered volatile by the presence of chlorides and 
fluorides in the magma and consequently are able to leave the ipiagnia 
and escape into the surrounding rocks. As they approach the Earth's 
surface they begin to react with the other fumarolic gases and are de- 
posited in the form of metallic minerals in the fissures through which 
the gases are streaming. Hematite is probably the commonest mineral 
formed in this way. During one of the eruptions of Vesuvius a fissure 
3 feet wide was thus filled with hematite in a few days. Galena, the 
chief ore mineral of lead, is occasionally formed at Vesuvius by the mu- 
tual action of the lead chloride and hydrogen sulphide contained in the 
fumaroles. Ores as it were are actually being deposited under our eyes; 
in fact, it was the contemplation of these phenomena on the flanks of 
Vesuvius that first suggested the fruitful idea that there is an intimate 
relation between igneous rocks and the occurrence and origin of ore 
deposits throughout the world. 

The temperatures of the gases issuing from fumaroles may be exceed- 
ingly high. In the remarkable fumarole field known as the Valley of 
Ten Thousand Smokes, near the volcano of Katmai in Alaska, tempera- 
tures as high as 645 C. have been measured. A view of several fuma- 
roles is seen in Fig. 183. m 


The volcanic cone near Naples known as Solfatara last erupted in 
1198; since then it has been merely discharging steam mingled with 
sulphur vapor, and this has given rise to the use of the term solfataric 
stage when volcanoes become quiescent or are dying. Some of the great 
cones of the Cascade Range, like Mt. Shasta, appear to be in a solfataric 
stage. In Yellowstone Park the solfataric condition still prevails and 
fumaroles abound. Although the steam given off in fumaroles can be 
mostly ascribed to magmatic origin, the amount is often increased 
by descending surface water that becomes vaporized, either by contact 
with hot rocks or by volcanic exhalations, which, as already mentioned, 

Fig. 183. General view of the Norris Geyser Basin, showing hot springs and fumaroles, 
and the white siliceous mass of geyserite deposited by them. Yellowstone Park. 

are chiefly superheated steam. This is probably the case in Yellowstone 

Carbon dioxide gas is given off copiously in many places where vol- 
canic activity still abounds, and in many where volcanism has long since 
died out. In some places the carbon dioxide issues directly from the 
ground as a gas spring, and such occurrences are known as mofettes. 
Being heavier than air, in still weather it may collect in depressions near 
the vent, and, as it is colorless, tasteless, and odorless, such pools of 
gas are deadly traps for animals by suffocating the creatures that enter 
them. This is illustrated by " Death Gulch" in Yellowstone Park, 
where animals as large as grizzly bears have become asphyxiated. But 
the carbon dioxide is far more likely to encounter ground water on its 
way upward and thus give rise to a carbonated spring, which if it passes 
through limestone will dissolve some of the limestone and is therefore 


likely to deposit carbonate at the Earth's surface. Some carbonated 
springs, however, derive their carbon dioxide from the action of acid on 


Hot springs as well as fumaroles are likely to occur in volcanic regions. 
There is an intimate relation between hot springs and fumaroles: in 
many regions as the dry season comes on some of the hot springs become 
fumaroles, and when the wet season returns the fumaroles become hot 
springs. The evident seasonal variation leads to the theory that hot 
springs are chiefly fed by ground water that has become heated by 
magmatic steam. The water circulation is thus in principle like the 
hot-water heating system in a house, but instead of a furnace in the 
basement supplying the thermal energy, magmatic steam furnishes the 

It is impossible in certain regions to tell how much of the water (and 
steam) of hot springs and fumaroles is of surface and how much is of 
magmatic origin. The amount of dilution probably varies in different 
regions. It is estimated that the hot springs at Lassen Peak consist 
of 10 per cent of magmatic water and 90 per cent of water of surface ori- 
gin. Hot springs in the rainless arid interior of some deserts have been 
regarded as mostly of magmatic origin. The proof that magmatic 
emanations have passed into such waters is found in the presence in 
them of such substances as arsenic, boric acid, and other constituents in 
quantities and under conditions that show that they could not have 
been leached out from the surrounding ro'cks of the country. In Yellow- 
stone Park it is probable that most of the water is of surface origin, 
which becomes heated in depth by the condensation of magmatic steam 
in it and returns in this heated condition to the Earth's surface. 

While there are various types of hot springs, according to temperature 
or substances in solution, the most interesting are boiling springs and 
geysers. Warm carbonated springs that deposit travertine have been 
already described in Chapter VI. 

Boiling Springs. Actively boiling springs are a feature of many 
volcanic regions. Many of them occur in Yellowstone Park, especially 
in 'the different geyser basins (Fig. 184). They grade from pools that 
are hot but rarely boil, or else simmer quietly, into springs that boil 
strongly and steadily, and even some that boil more or less violently 
and with somewhat explosive energy, interrupted by short periods of 
repose. The latter form transitions to the geysers mentioned beyond. 
So long as the supply of water is sufficient to enable the spring to have 
an overflow it remains limpid, and it usually has a deep blue or green 


color, but if the evaporation through boiling is equal to the inflow, the 
water is more or less turbid from particles of disintegrated rock, and 
eventually becomes a mass of boiling mud. The mud may be white, or 
variously tinted yellow, red, purplish, or black by oxides of iron and 
manganese, and such hot springs are called " paint pots," " mud pots," 
etc. The mud as it increases in amount becomes so thick and viscid 
as to prevent regular ebullition, and, owing to the accumulating steam 
pressure, the paint pot boils spasmodically and with some violence, the 
mud being thrown into the air and about the vent, where it collects in 

Fig. 184. "The Devil's Punch Bowl." A hot spring boiling in the cup-like deposit 
of geyserite it has formed. The opening is several feet in diameter. Upper Basin, Yel- 
lowstone Park. (U. S. Geol. Surv.) 

considerable masses. These are known as mud volcanoes, or mud gey- 
sers. They usually mark a declining stage of activity in the life of a 
hot spring. 

Geysers. This term, from an Icelandic word meaning to gush, is 
applied to certain hot springs that at intervals spout a column of hot 
water and steam into the air. Depending on the size of the geyser and 
its special peculiarities, the height to which the column of water is 
ejected ranges from only a few feet up to several hundreds; the eruption 
may last a few minutes or several hours; the quantity of water dis- 
charged may be small or it may be many thousands of gallons; the- jet 
may play steadily and continuously straight up, or it may be fitful, be 
composed of minor jets, or be thrown in inclined directions. The 
interval between eruptions may be a definite number of minutes or 
hours, or it may be irregular, and several clays may elapse between erup- 
tions. Each geyser has in these ways its own peculiarities. As they 
are boiling springs of a special kind they are not common and are 



almost wholly confined to three principal regions: Yellowstone Park, 
Iceland; and New Zealand. 

Some geysers consist at the surface of a basin, which may be several 
feet to a number of yards across, and rather deep. The sides and edges 
of the basins usually are beautifully ornamented by the deposits of silica 
described beyond, and they terminate at the bottom in tubes or fissures 
leading to the heated depths below, as shown in the diagram, Fig. 186. 
The tubes and basins are, except after eruptions, filled with water at or 
near the boiling point. In other types the geysers by their deposits 
have built up mounds, or cones, of silica, from a foot or two to several 

Fig. 185. Lone Star Geyser in eruption, showing cone of geyserite. Yellowstone 

Park. (Haynes.) 

yards high, which form upward continuations of the pipes (Fig. 185). 
Of the Yellowstone geysers the most celebrated perhaps is the one known 
as " Old Faithful," which for many years after its discovery had a very 
regular interval between eruptions of about 65 minutes. It is now less 
regular, ranging from 60 to 80 minutes. This, and the decline of activity 
in other geysers, or springs, does not mean any immediate diminution 
of thermal action in this region, but only changes going on in the under- 
ground system of pipes and fissures that supply the hot water. Alto- 
gether there are several dozen fine geysers in the park, and the number 
of hot springs, fumaroles, and thermal vents of various kinds amounts 
to several thousand. 



Cause of Geyser Action. The intermittent eruptive action of geysers 
depends on the relation between pressure and the boiling point of 
water, as was pointed out by Bunsen in connection with the great 
geyser in Iceland. The boiling point of water under the ordinary 

pressure of the atmosphere at sea level is 
212 F.; increase of pressure raises it, a 
decrease lowers it. Thus the boiling point 
at the bottom of a column of water will be 
raised by the pressure of the superincum- 
bent column above it; as shown in Fig. 186, 
it will gradually rise as we follow the tube 
from the surface downward. If, however, 
the cavity or fissure is large and open, the 
heated water below will rise, convection 
currents will be established, mixing the 
water, so that it will have nearly, though 
not quite, the same temperature in differ- 
ent parts of the cavity, and a regular boil- 
ing spring will result. But if the tube is 
long, narrow, tortuous, or constricted, con- 
vection will be prevented or restrained, and 
the water must boil in different levels at 
different temperatures corresponding to the 
pressures. Suppose at a point 230 in the 
Fig. 186. Vertical section to figure the boiling point is reached; bubbles 

illustrate conditions necessary for Q f steam are formed, the Column of water 
geyser action. . ' 

above is raised a little by the expansion, 

the bubbles of steam rise in the cooler liquid above and collapse, the 
column of water settles back with jarring, thudding sounds commonly 
heard before eruption. The temperature of the water will gradually 
rise until it is just about at the boiling point for each level corffespond- 
ing to its depth and pressure. Finally when a sufficient volume of steam 
is formed in the lower part of the geyser tube, the expansion will cause 
some of the water in the basin or cone at the top to overflow. This 
overflow lowers the pressure throughout the tube, and the water at 
each level, being now heated above the boiling point for the diminished 
pressure, will immediately flash into steam, and a mingled column of 
steam and hot water will be driven roaring out of the pipe into the air. 
After the eruption is over, the system fills again by inifow of ground 
water into the geyser tube, and the process is repeated. 

The varied forms of fissures, underground conduits, and water supply 
account for the peculiarities shown by different geysers. It was found 


by accident that adding alkaline substances, such as soap or lye, to the 
waters of geysers causes some of them to erupt very quickly. The 
government had to put a rigorous ban on the " soaping " of the geysers 
of the Yellowstone National Park, in order to prevent mining them. 

That the source of heat for the geysers and hot springs in the Yellow- 
stone Park is deep-seated is shown by their occurrence in and on the 
shores of Yellowstone Lake, an immense body of very cold water, be- 
neath which the rocks must be cooled to a considerable depth. 

Hot-Spring Deposits. It has been previously shown that warm 
springs, especially if they contain carbon dioxide in notable quantity 
and come up through limestone beds, form deposits of calcareous tufa, or 
travertine. (See Chapter VI.) But the waters of boiling springs and 
geysers, which occur only in regions of recent volcanic activity, are mostly 
alkaline and carry silica (SiO 2 ) in solution, which they deposit as a 
whitish material. This deposit ranges from compact to spongy in tex- 
ture, and is known as geyserite, or more commonly as siliceous sinter. 
It forms the geyser cones, or is deposited as incrustations, much of it of 
great beauty, in and about the margins of the hot-spring and geyser 
basins. The geyser waters are dilute, in fact, so dilute as to be tasteless, 
and the rate of deposition is very slow when it occurs only through 
evaporation but is hastened by the action of organisms. Deposits of 
considerable size and thickness have been, and are being, made in this 
way, as seen forming the floor of the basin in Fig. 183. While hot 
springs and geysers are not geological factors that are important be- 
cause of the results they achieve, nevertheless they are of great sig- 
nificance in a proper understanding of certain processes, such as the 
deposition of some ores of metals; and furthermore they are of wide 
popular interest. 

As in the case of travertine, the deposition of silica is largely due to 
its secretion by low forms of vegetable life (diatoms and algae, the latter 
related to seaweeds), which flourish in the warm waters and even in the 
hot waters. The beauty of many of the pools is greatly enhanced by 
the rich coloring that these growths give to them. 

Besides silica, the hot springs deposit other substances. The waters 
of some springs are acid and deposit sulphur and alum salts; and from 
other springs sulphides of arsenic and of metals are deposited, thereby 
throwing light on the processes by which ore bodies are formed. 

In recent years it has become apparent that fumarole fields may be 
developed so as to yield large supplies of steam for power generation. 
The fumarole field in the volcanic area of Tuscany, north of Rome, was 


the first to be developed. Wells have been put down to depths of 600 
feet, and the flow of steam, as well as the temperature of the steam, has 
been found to increase as depth is gained. About 30,000 horsepower is 
generated and is transmitted to Florence, 60 miles away, Pisa, and other 

At " The Geysers," in the Coast Ranges of California, 40 miles north 
of San Francisco, is an area of 35 acres containing a few feeble fumaroles 
and some small but very hot springs. The name " Geysers " is a mis- 
nomer, however, as none of the hot springs is periodically eruptive. 
In 1921 the idea was conceived of drilling wells in this area to develop 
a flow of steam for power purposes. So far eight wells have been put 
down, the deepest being 650 feet, and copious supplies of superheated 
steam have been developed. In fact, it is estimated that four of the 
wells will on the average deliver more than 1300 horsepower each. As 
in Tuscany the deeper a well is drilled the greater the steam flow and 
the hotter the steam. 

In Java also the power possibilities of its fumarole fields are being 
investigated. One of the fields was bored in 1926. The most promising 
well, 220 feet deep, yields steam sufficient to generate 1200 horsepower. 
Many fumarole fields remain to be bored in Java, as well as others in the 
nearby islands of Sumatra and the Celebes. 


1. Volcanoes: Their Structure and Significance; by T. G. Bonney. 3rd edition. 
379 pages. G. P. Putnam's Sons, New York, 1913. 

Accurate, readable, fairly up-to-date book on volcanoes. 

2. Hawaii and its Volcanoes; by C. H. Hitchcock. 314 pages. The Hawaiian 
Gazette Co., Honolulu, 1909. 

A lucid description of the Hawaiian volcanoes, especially Kilauea and Mauna Loa. 

3. Tolcanoes of North America; by I. C. Russell. 346 pages. The Macmillan 
Co., New York, 1897; reprinted 1924. 

An interesting, well-written account of the volcanoes of North America, but un- 
fortunately somewhat out of date. 

4. Der Vulkanismus; by F. von Wolff. I. Band: Allgemeiner Teil, 711 pages, 
1913-1914. II. Band: Spezieller Teil, 1923 (still in progress). Ferdinand Enke, 

Technical; exhaustive, with ample bibliographies. 

5. Vulkankunde; by Karl Sapper. 424 pages. J. Engelhorns Nachf ., Stuttgart, 

The best general book on volcanoes. 


The igneous rocks form one of the great divisions of the rocks that 
make up the crust of our planet. As- their name implies, heat was an 
essential factor in their origin, and they may be defined as those rocks 
which have been formed by the solidification of molten matter that originated 
within the Earth. Such molten matter, as was explained in the discussion 
of volcanic action, is commonly called magma, a term we shall frequently 

Distinguishing Characters. The features that distinguish the 
igneous from the sedimentary and metamorphic rocks consist partly in 
the relation that the igneous masses exhibit towards other rocks with 
which they occur or with which they are in contact (a relation that we 
term their mode of occurrence), and partly in the characters that be- 
come evident when the rock itself, instead of the mass of which it is a 
part, is closely examined. 

Igneous rocks do not of course contain fossils, nor do they show as 
a rule the parallel or banded appearance of the stratified rocks. They 
have also certain distinctive peculiarities in the arrangement of their 
component mineral grains (the texture, as it is called). Some igneous 
rocks, indeed, are more or less made of glass, which at once betrays 
their origin, because glass is formed only by the chilling of molten ma- 
terial. The textures will be described when the different kinds of ig- 
neous rocks and their classification are considered; first we will discuss 
the ways in which masses of igneous rocks occur as elements in the archi- 
tecture of the Earth's crust. 


Intrusive and Extrusive Rocks. There are two chief modes of 
occurrence of igneous rocks: the intrusive and the extrusive. In the 
intrusive mode of occurrence the magma, at the time it was rising from 
the depths, stopped before reaching the surface, and consequently it 
cooled and solidified under the cover of the rock masses of the Earth's 
outer shell. In the extrusive mode of occurrence, the magma attained 
the surface: it was extruded upon it, and has solidified there. The 
extrusive rocks are sometimes called effusive and sometimes volcanic 



rocks, although the term volcanic is perhaps not strictly applicable to 
those lavas that were extruded from fissures rather than from volcanic 

Although the division of igneous rocks into intrusive and extrusive 
is a natural one, the two classes are closely connected and, in fact, grade 
from one into the other. The magma of any extrusive mass came up 
through some passageway from below; this passageway remained filled, 
and eventually the magma in it solidified into rock. Consequently in 
theory every extrusive body that occurs on the Earth's surface is con- 
nected with an intrusive mass occurring below. In some places this 
connection of the extrusive mass with its root may be seen, but more 
generally the extrusive covers the root or has been separated from it by 
erosion, and the former continuity has been destroyed. It is clear also 
that we must think of the intrusive prolongation as extending downward 
and connecting with some greater mass of magma (or rock) below, of a 
nature to be presently described. See in this connection what has been 
said regarding the relation between volcanoes and deep-seated masses 
of magma in the preceding chapter. 

Both intrusive and extrusive rocks have various modes of occurrence. 
The mode of occurrence of intrusive masses depends on how the intrusive 
mass is related to the rocks that enclose it; and the mode of occurrence 
of extrusive masses depends on the conditions under which the magma 
was ejected. We shall begin with the modes of occurrence of intrusive 
masses; but it should first be recalled that, inasmuch as these intrusive 
masses were covered at the time of their formation by the rocks into 
which they were intruded, they can be exposed at the Earth's surface 
and thus laid open to observation only after erosion has carried away 
the cover and has disclosed their intrusive character. In some places, 
where the magmas were intruded under a thin cover, the time necessary 
to do this may have been short; but in other places, where the igneous 
masses were deeply buried, it may have been exceedingly long. 


The principal kinds of intrusive igneous bodies are dikes, sills, lacco- 
liths, necks, stocks, and batholiths. Several other modes of occurrence 
are recognized, but as they have not the importance of those mentioned, 
they will be treated for simplicity's sake as modifications of them. The 
simplest form of intrusive body is the dike and this will be considered 

Dikes. A dike results from the solidification of magma that has 
filled a fissure in preexistent rocks. Consequently, it is a tabular mass: 



its length and breadth are great compared with Its thickness. It may 
" cut," that is, pass through, rocks of any kind igneous, sedimentary, 
or metamorphic. In the sedimentary rocks it must by definition cut 
the planes of stratification at an angle; if, however, the igneous mass 
lies parallel to the bedding planes it is termed a sill. A dike may be a 
few yards or many miles long; it 
may be a fraction of an inch, or 
many hundreds, or even some thou- 
sands of feet thick. An illustration 
of a dike is seen in Fig. 187. 

Most dikes are from 2 or 3 feet 
up to 20 thick; the length varies 
greatly. A great dike in the north 
of England extends for over 100 
miles. The angle of inclination of 
the plane of extension of the dike 
with the horizontal is called its dip. 
The direction of its outcrop, or in- 
tersection with the horizontal plane, 
is termed its strike, or trend. 

Some dikes have attained the 
Earth's surface and given rise to out- 
flowings of. lava, but others have 
not reached it and have become ex- 
posed by subsequent erosion. Some 
dikes were the canals that fed larger 

intrusive bodies above them, such as the sheets and laccoliths to be next 
described. In the process of erosion, a dike may be more resistant than 
the surrounding rocks and hence is left projecting as a wall; some, 
however, are less resistant and form ditches; from these features the 
name is derived, especially from its resemblance to the more prominent 
wall, for dike means both wall and ditch. The rock of some dikes is 
divided into blocks by joints, and very commonly the blocks are columns 
lying perpendicular to the walls of the dike, like a pile of cordwood, an 
arrangement whose origin is described later under columnar structure. 
Dikes occur at many places in more or less well-defined systems, and 
around volcanic centers are likely to be radially disposed. 

Sills. It is not uncommon to find, where magma has been intruded 
in bedded rocks, that it was injected as layers between the beds. In- 
jection of this kind most frequently happens where the beds are easily 
penetrated, as in shales, thinly bedded sandstones, and the like. An 
igneous mass that thus lies concordantly between the bedding planes is 

Fig. 187. Dike of trap rock in granite. 
This dike is less resistant than the enclos- 
ing granite and has been cut away by 
erosion, leaving a trench in the granite. 
Isles of Shoals, N. H. 



known as a sill, or an intrusive sheet. Sills may be a foot or so in thick- 
ness or several hundred feet, and they may be many square miles in area. 
An illustration of them is seen in Fig. 188. 

Sills may break dike-like across the strata and then continue along a 
new horizon. Sills may be distinguished from flows of lava that have 
been buried by deposits of later sediments by the fact that the rock com- 
posing the sills is of the same compact nature at top and bottom, i.e., 
is not slaggy or scoriaceous like a lava flow, and by the overlying sedi- 

Fig. 188. Sills of Igneous rock. Cottonwood Canyon, New Mexico. 
(U. S. Geol. Surv.)' 

ments having been baked and altered by the intrusion. The surface of 
a lava flow is usually spongy, ropy, slaggy, etc., and a flow could of course 
exert no action on beds not yet deposited upon it. Sills are most likely 
to occur where larger intrusions of magmas, such as laccoliths and stocks, 
have taken place, as accompanying features in the surrounding strata. 
In regions where thick sills occur and the strata have been dislocated and 
upturned, they may give rise. to prominent topographic and scenic land 
features through the effects of later erosion. This is illustrated in some 
of the trap ridges of southern New England, northern New Jersey, and 
in other places. 

Laccoliths. The laccolith in its typical development is a lenticular 
or dome-shaped mass of igneous rock that was intruded into sedimentary 
rocks between the bedding planes. It has a flat floor, and is more or less 
circular in ground plan. If during the formation of a sill, magma is 
supplied more rapidly from below than can easily spread laterally 



away from the supply channel, the overlying strata will be arched up, 
as if by a hydraulic press, and a thick lens of liquid rock will be pro- 
duced, giving rise on solidification to a laccolith. Such a mass may be 
a few hundreds of feet or more than a mile thick at the center, and a few 

Fig. 189. Section of a laccolith. The black area is the igneous rock. 

hundreds of yards or many miles in diameter. A section of a laccolith is 
shown in Fig. 189, and a photograph of one from which the cover has 
been removed by erosion, exposing the igneous rock, is seen in Fig. 190. 
While the above statement gives the idea of a typical laccolith, many 
departures from this arrangement are found in the actual occurrences. 

Fig. 190. Bear Butte, a laccolith denuded of its cover and the igneous mass laid bare. 
The ring of upturned eroded strata is seen about its base. Black Hills, South Dakota. 
(U. S. Geol. Surv.) 

In ground plan they may be circular, oval, or quite irregular, and in- 
stead of being symmetrical in section, as in Fig. 189, they may be wedge- 
shaped. ^ According to their degree of flatness, all transitions into sills 
occur. They may also break across the strata in places like sills. They 
may thin out into sills, or be accompanied by sills on the flanks of the 
arches, and thus be compound in structure. Such sills may themselves 
swell out into inclined lenticular masses, or subordinate laccoliths. 



And in regions where strata were being folded, areas of relief from pres- 
sure or openings might form on the sides of the arches that would permit 
the entrance of magma (Fig. 191). This would give rise to inclined, 
doubly convex bodies like that shown in Fig. 192. Laccolithic bodies 
of this character have been termed phacoliths (from the Greek words 
for lentil + stone). 

It is sometimes asked whether the magma itself supplies the force 
necessary to make room for itself by arching up the strata; in other 

Fig. 191. Strata being folded by com- Fig. 192. Section of an inclined lacco- 
pression CD, relief from pressure from lith, or phacolith. 

overlying beds and spreading might occur 
in direction AB* 

words, was the magma aggressivej or as indicated above, was the force 
lifting the beds produced in some other way, and did the magma simply 
flow into the space that was opening for it, the intrusion having been, 
so to speak, permissive? A study of known laccoliths shows that in all 
probability both of these modes of intrusion occur. In central Montana, 
where the strata are horizontal and undisturbed save where intrusive 
masses occur, the magma must have acted aggressively, but in other 
places where folding and uplifts occur, the intrusions were probably per- 

Laccoliths, more or less exposed by erosion, are conspicuous features 
in many parts of western North America, where they were first discov- 
ered by Gilbert in the Henry Mountains of southern Utah. Subse- 
quently they have been found in various parts of the world and are there- 
fore a not uncommon form of intrusion. Moreover, some of the more 
recently recognized laccoliths are of immensely greater size than the 
classic laccoliths of the Henry Mountains, the largest of which has a 
volume of 10 cubic miles. The great Duluth laccolith on the northwest 
shore of Lake Superior is estimated to represent the injection of 50,000 
cubic miles of magma. 

The floor of a large laccolith as a rule is not horizontal but dips from 
the perimeter inward toward a point under its center, just as if the floor 
had sagged from loss of support. Such loss of support is conceivably the 
result of the emptying of a magma reservoir below the laccolith. 

Necks. When a volcano becomes extinct, the column of magma that 
filled the conduit leading to unknown depths below will solidify and form 



a cylindrical mass of igneous rock. Erosion will in time cut away a 
great part of the ashes and lavas of the cone, leaving this more compact 
and resistant rock projecting, as shown by the line abc in Fig. 180." 
The level of erosion may eventually descend into the rocks that form the 
basement on which the volcano stands; and all of the ashes and lavas 
having been thus swept away, only the projecting mass remains to mark 
the former site of the volcano. Such a mass of rock is known as a 
volcanic neck (Fig. 193). It is commonly more or less circular in ground 

Fig. 193. Alesna volcanic neck, Mt. Taylor region, New Mexico. (U. S. Geol. Sunr.) 

plan and may be from a few hundred yards up to a mile or more in 
diameter. The rocks about volcanic necks are likely to be cut by a 
radiating system of dikes, and commonly, if stratified, injected with 
sills. The significance of volcanic necks has been previously explained 
in the discussion of volcanoes. 

Stocks. This term is applied to certain large domal bodies of in- 
trusive rock that in the form of magma have ascended into the upper 
levels of the Earth's crust and there solidified. They have become vis- 
ible because erosion has stripped off the covering rocks. They have 
as a rule a more or less circular or oval ground plan. Their outer sur- 
face, or contact surface, cuts across the inclosing rocks, is more or less 
irregular, and the mass may widen in extent as it descends. Their size 
ranges from a few hundred yards to several miles in diameter. As they 
are likely to form protuberant topographic features after being exposed 
by erosion, they are sometimes called bosses. The distinction from a 



volcanic neck is not one of size alone, though necks tend to be smaller 
than stocks, but in that the term " neck " is employed only when there 
is evidence that the igneous body has functioned as the supply conduit 
of a volcano. Some stocks were doubtless necks, but this cannot now 
be proved. 

Batholiths. A batholith is a huge intrusive mass of igneous rock. 
According to some definitions, a batholith is " bottomless/' that is, it 
extends indefinitely downward into the Earth's crust, in contrast to a 
laccolith, which rests on a floor. The " bottomlessness " is of course 
pure hypothesis. A batholith differs from a stock only in its much 
greater size, as some are exposed over many thousands of square miles 
of surface. Arbitrarily, an intrusive igneous body less than 40 square 
miles (100 square kilometers) in areal extent is called a stock; if larger 
than 40 square miles, a batholith Some stocks, as shown in Fig. 194, 

Fig. 194. Diagram to illustrate the occurrence of igneous rocks: b, batholith, partly 
uncovered by erosion; s, stock, partly uncovered by erosion; n, volcanic neck forming v, 
a volcano with tuffs and breccias; Z, Z, laccoliths; i, intrusive sheet, or sill; e, extrusive 
sheet; d t d, dikes. The batholith and stock belong to an older generation than the other 
igneous bodies shown. If the section went deep enough, it might show a younger batho- 
lith from which the smaller bodies and the volcanic materials were derived. Horizontal 
distance shown thirty miles; vertical distance, three miles. 

are merely dome-like protuberances from the body of an underlying 
batholith; they have been aptly called cupolas. 

The largest batholith in the United States is the Idaho batholith in 
central Idaho, extending over an area of 16,000 square miles. The 
causes for the rise of so stupendous a mass of molten rock matter into 
the higher levels of the Earth's crust and the processes by which it makes 
room for itself are among the most fundamental and fascinating problems 
of Geology. 


The modes of occurrences of the extrusive igneous rocks have already 
been described in connection with volcanoes and extrusions of lava, and 
what has there been said in regard to them may be profitably consulted 
in this connection. For the sake of convenience the following summary 
is here given. Magma is erupted in two ways, depending on the quan- 
tity and activity of the gases contained in it: the quiet, in which it wells 


out as a liquid and solidifies into rock, and the explosive, in which it is 
violently driven into the air and falls in the form of solid fragments. 

Quiet Eruption: Lava Flows. Magma that reaches the surface and 
pours out is known as lava. When solidified it is commonly spoken of 
as a lava flow, or an extrusive sheet. Usually such flows are poured 
out from volcanoes. The extrusions of a few volcanoes like some of 
those in Hawaii, are indeed almost wholly of this nature, but generally 
lava flows succeed or alternate with ejections of fragmental material. 

Some lava flows have not been erupted from volcanoes, but have 
welled out quietly from fissures. In the geologic past fissure eruptions 
have occurred on a huge scale, as in the Columbia River region of the 
northwestern United States, in western India, and in the north of the 
British Isles. In each of the first two regions the pile of superposed lava 
flows is thousands of feet thick and covers an area of approximately 
200,000 square miles. 

Not infrequently sheets of lava have sunk below sea level and been 
covered by deposits or they have been erupted on the sea floor and have 
been covered by sedimentation. Still more commonly lava flows on 
the land have been buried under various kinds of continental deposits. 
Such buried extrusive sheets are distinguished from sills by the fact 
that they have not altered or baked the sediments above them, and their 
upper surfaces usually show the structure common to the surface of 
lavas, such as the vesicular, scoriaceous, and ropy ones described previ- 
ously. Furthermore the layer of sediments directly above a buried 
lava flow is likely to contain pebbles or boulders derived from the upper 
part of the flow before it was entirely buried. 

Explosive Eruption: Tuffs and Breccias. When magma attains 
the Earth's surface in the conduit of a volcano, it may erupt as quiet 
flows of lavas as already mentioned, or, if its viscosity is sufficient and 
it is charged with vapors under great tension, it will give rise to explosive 
activity, and the material will be hurled into the air. Owing to the 
expansion of the contained gases, chiefly steam, the ejected pieces 
usually are more or less vesicular, and range in size from large masses 
to fine dust. This material is roughly classified according to size, as 
previously explained. 

During an explosion in a volcanic vent not only are fragments of hot, 
still-fluid magma ejected, but also great quantities of cold solidified 
lava disrupted from the crater walls are blown into the air. The coarse 
angular pieces produced by the fragmentation of the cold lavas are 
termed blocks, in centra-distinction to the bombs, whose roundish forms 
show that they were still viscous during their aerial flight. The coarser 
material the blocks, bombs, ashes, and lapilli falls around the vent 


and builds up the cone; the lighter ashes and dust, carried by air 
currents, tend to fall after these, and at greater distances. The coarser 
material thus produced is termed volcanic breccia, while the finer ma- 
terial is known as tuff. 

Tuffs and breccias are widely distributed, occurring wherever volcanic 
activity is being or has been displayed, and their presence is, indeed, one 
of the surest indications of former volcanism in places where it has long 
since died out. We are thus able to recognize that volcanoes formerly 
existed in various parts of the eastern United States and Canada, from 
Nova Scotia to Georgia. Tuffs and breccias occur in vast quantities 
piled up in places thousands of feet in thickness in the Rocky Mountains, 
where, as in western Wyoming, serried mountain peaks have been sculp- 
tured from them by erosion. 


The geologic period when a given mass of igneous rock was erupted 
or intruded is determined by ascertaining its .relations to the rocks with 
which it has come in contact'. Thus, if a body of igneous rock, such as a 
dike, cuts through another body of rock, it is manifestly the younger of 
the two. If it cuts across stratified beds it is younger than they are, and 
lavas of course are more recent than the rocks upon which they lie. 
If a sheet of igneous rock lying concordantly between strata has affected 
the beds both above and below it (see contact metamorphism, page 352), 
it is younger than both. If it has not baked or otherwise metamor- 
phosed the overlying beds, the sheet may be a lava flow and older than 
they are. It is thus usually easy to tell when an igneous mass is younger 
than other rocks by examining its contacts with them. The age of the 
stratified rocks is of course determined by the fossils that occur in them, 
and the endeavor is made to find whether the igneous rocks, which con- 
tain no fossils, are older or younger than the fossiliferous rocks with 
which they may be in contact. 

Introductory. The features by which the igneous rocks are dis- 
tinguished have been already mentioned in a preliminary way in dis- 
cussing the products of volcanoes, but they should now receive the 
attention that they demand. Igneous rocks are divided into different 
kinds on the basis of two properties: first, their texture; and second, 
their composition. Each of these properties requires explanation. 

Texture. The most obvious thing about an igneous rock is its 
texture. By texture is meant the relative size or sizes of the component 


grains and the shape and arrangement of the grains. Thus if the 
grains are as large as peas, we say that such a rock is coarse-grained 
in texture; if the grains are the size of those in granulated sugar, we say 
that the rock is fine-graimd; whereas, if the particles are so minute 
that they cannot be discriminated by the unaided eye and the rock looks 
as if it were a homogeneous substance, we say that it is aphanitic 1 in 

The texture depends on the rate at which the magma cooled. For, if 
the magma is too hot, as previously explained, crystallization cannot 
take place, and no crystals will begin to form until the temperature has 
fallen far enough; then they will begin to separate from the magma and, 
if the cooling is very slow, they will have time to grow to large size, 
thus producing a coarse-grained rock. But, if the cooling is rapid, more 
and more new centers of crystallization will be forced to form, and, if 
the process is thus hurried, instead of few crystals growing to large sizes, 
the rock will consist of a large number of smaller particles and will, 
therefore, be fine-grained in texture. And with still more rapid cooling 
the particles may be so minute that they are not discriminable by the 
unaided eye and the resultant rock is of aphanitic texture. Analogy will 
now carry us one step more: we can conceive that the cooling may take 
place with such great rapidity that the magma will solidify into a 
homogeneous substance before any crystallization, which consists in the 
molecules arranging themselves together to form definite solid com- 
pounds, can occur. In this event the result will be a glass, or a glassy 
rather than a stony texture, a case that is by no means uncommon. 

To sum up, then, we see that igneous rocks may be coarse-grained, 
fine-grained, aphanitic, or glassy in texture, and that which of these 
textures is developed depends on the rate of cooling of the magma. 

Porphyry: Porphyritic Texture. In what has been said so far re- 
garding the texture of igneous rocks, it has been tacitly assumed that 
the component mineral grains in any given rock are of uniform size, or 
that the rock is evenly granular, as it is called. Not all igneous rocks, 
however, are evenly granular. Inspection shows that many of them 
are composed of crystals of two sizes: some crystals that are larger and 
more distinct and which are embedded in a matrix of much finer grains. 
An igneous rock having this texture is called a porphyry. Examples of 
the even-granular and porphyritic textures are seen in Fig. 195, A and B. 
The matrix of a porphyry is termed the groundmass, and the large crystals 

1 Dense is often incorrectly used in America as a synonym for aphanitic. It is 
also used correctly to mean " of high density " and this double usage leads to am- 
biguity. A "dense" felsite (in the sense of an ultra-fine grained rock) is not a 
dense rock. 



embedded in the groundmass are called the phenocrysts (clearly dis- 
cernible crystals). Porphyritic rocks are common. 

The groundmass may itself vary widely in grain size in different 
porphyries: it may be medium-grained, fine-grained, aphanitic, or 
glassy; most commonly it is aphanitic, as in the extrusive rocks. The 
phenocrysts also may vary widely: they may be of large size, as large 
as walnuts or as small as grains of sand; they may be abundant or 
comparatively few. But in all porphyries there is this contrast between 
sizes of crystals, between groundmass and phenocrysts, which makes the 
essence of a porphyry. The student should guard against thinking that 
the porphyritic texture is a contrast of colors; thus a rock consisting of 

I* \ 

f** M i% ^i't f *f''t ' 


.,, ,,i .JKOf.^. -.Ai ^|.^' 

" ; "'^fi : 

:> _, i 

Fig. 195. A. Even-granular Rock. 

B. Porphyry. 

grains of light-colored quartz and feldspar, in which are embedded a 
few black crystals of mica, all grains being of about the same size, is not 
a porphyry. 

Relation of Texture to Geologic Mode of Occurrence. Since, as 
has been shown, the texture of an igneous rock depends chiefly on the 
rate at which the magma cools, it is clear that this rate will depend in 
turn most largely on the volume of the magma. Obviously an intrusive 
mass of magma that is surrounded and blanketed above by other 3 older 
rock masses must lose heat much more slowly than an extrusive one, 
which is poured out on the surface in the form of lava. Hence, as a 
coarse-grained texture is the result of slow cooling, we naturally asso- 
ciate it with intrusive masses, and, conversely, we regard the aphanitic 
or glassy textures as belonging to the products of extrusion the 


lava flows. But it is also clear that the size of the intruded mass will 
greatly influence the rate of cooling, since a very large mass cools more 
slowly than a small one. Thus, the rocks "that make up great batho- 
liths are coarse in texture, whereas the rocks of dikes and sills tend to 
be much finer-grained. On the other hand, the central portion of an ex- 
tremely thick lava flow may cool with sufficient slowness to develop a 
medium-grained texture, whereas a magma that was forced into a narrow 
fissure in cold rocks might be chilled so quickly as to assume an aphanitic, 
or even glassy texture. Thus various modifications of the general rule 
according to particular cases can be easily imagined; nevertheless this 
general rule, that the intrusive rocks are medium- to coarse-grained 
and the extrusive rocks are fine-grained to aphanitic, holds true. 

An important deduction that follows from the above is that the 
coarse-grained rocks, because they have been formed deep within the 
crust, can become visible at the Earth's surface only after a period of 
denudation that has been sufficiently long to remove the covering rocks 
and expose the igneous mass. 

While the rate of cooling is the most important factor that determines 
the texture of igneous rocks, as discussed above, it is not the only one. 
The subject is too complex for detailed treatment in this work, but it 
may be mentioned that the chemical composition also has its influence. 
Under similar conditions of cooling, basic magmas (those low in silica 
and high in iron and magnesia) tend to assume a coarser grain than those 
composed of much silica, alumina, and alkalies. The reason is that the 
basic magmas are more fluid than the siliceous magmas, as already ex- 
plained, and thus when they crystallize this mobility of the molecules 
permits the crystal grains to grow to larger sizes. 

Also the presence of the included gases that magmas contain, espe- 
cially the water, increases the fluidity and thereby promotes a coarser 
crystallization. This is very notably shown in and around certain in- 
trusive masses by dikes that are made up of large and even huge crystals 
of quartz, feldspar, and mica. Crystals several feet in diameter are not 
uncommon. The very coarse masses of this composition are known as 
granite pegmatites, and from them are obtained the plates of mica that 
are used commercially. In allusion to their giant grain the pegmatites 
are sometimes called giant granites. 

In a volcanic neck the rock is likely to be comparatively coarse- 
grained in spite of its small mass, because the constant upward passage 
of molten material to the surface causes the rocks surrounding the con- 
duit to become greatly heated, thus producing slow cooling of the last 
charge of magma that occupied the conduit and solidified there when the 
volcano became extinct. 


Composition of Igneous Rocks. Since igneous rocks are produced 
by the solidification of magmas, their composition will obviously depend 
on the chemical composition of these molten fluids. It has been already 
shown (Chapter X) that a magma consists of two parts: a volatile part, 
consisting of the fugitive constituents water vapor, carbon dioxide, 
sulphur fumes, amounting to a few per cent, and a non-volatile part, 
consisting of the fixed constituents, chiefly molten silicates. Although 
for reasons we cannot now consider, the gaseous part is important in 
rock formation, it is essentially the molten silicates that give rise to the 
rocks and are the chief constituents of magmas. While it is evident 
that a magma cannot be analyzed directly, still the fixed constituents 
and their relative proportions can be ascertained by analysis of the cold 
and solid rock. Many thousands of igneous rocks from all parts of the 
world have been analyzed and the following results have been obtained, 
which show the ranges in the amounts of the various constituents. 
These results are reported by the chemist in terms of oxides, and the 
convention has grown up in speaking of the chemical composition of 
magmas and rocks as if they were actually composed of these oxides. 
When we say that a rock is high in magnesia we really mean that its 
chemical analysis shows that a large amount of magnesia is present: 
this mode of expression implies nothing as to the chemical combination 
of the elements in the rock or magma. 

Silica, SiO2, always present; ranges from 35 to 80 per cent. 

Alumina, AI 2 O 3 , ranges from to 25 per cent. 

Oxides of iron, FeO and Fe 2 Os 3 usually both; to 20 per cent. 

Magnesia, MgO, to 45 per cent. 

Lime 7 CaO, to 20 per cent. 

Soda, NaaO, to 16 per cent. 

Potash, K 2 O, to 12 per cent. 

It must not be concluded from the above table that any and all sorts 
of mixtures of these oxides can occur within the limits shown. As we 
shall see presently, certain general laws govern their associations. It 
will also be noticed that there is one acid-forming oxide (silica) present, 
while the oxides of the six metals (aluminum, iron, magnesium, calcium, 
sodium, and potassium) are in general bases. Oxides of other elements 
occur in small or minute quantities, but are of so much less importance 
that they may be neglected. 

Associations of Oxides in Rocks. Although there are many exceptions 
to this rule, it is generally true that large percentages of potash and soda 
(alkalies) in a rock are accompanied by a correspondingly large amount 
of silica and consequently by small amounts of the other three metallic 
oxides. Conversely, large percentages of magnesia, lime, and iron ox- 


ides are likely to be associated, and these go with low silica, the alkalies 
being small or wanting. These reciprocal relations are of fundamental 
importance in igneous rocks, and it will be recalled that they have been 
pointed out before, because the nature of volcanic activity and the kinds 
of volcanoes in large measure depend on them. They may be expressed 
in a general way as follows: 

Where Si0 2 , (Na ; K) 2 are high, CaO, MgO, FeO are low or wanting. 
Where CaO, MgO, FeO are high, Si0 2 is low, and (Na,K) 2 are low 
or wanting. 

Crystallization. It is a familiar experiment that if a liquid contain- 
ing a salt, zinc sulphate for example, is boiled down and concentrated to 
a certain point, all of the zinc sulphate can no longer remain in solution, 
but will begin to appear as a solid in the form of crystals. If the hot 
solution is allowed to cool, more crystals of the salt will be formed, since 
hot solutions as a rule can contain more salt than cold ones. In analogy 
with this, a"magma can be regarded as a solution; it contains dissolved 
in it various salts (mineral molecules) more or less electrolytically dis- 
sociated. If the magma cools with sufficient slowness, the dissolved 
matter in it will separate from it as crystals. This crystallization will 
proceed as the temperature falls until the whole magma has turned into 
a mass of solid crystal grains. The molten liquid has become stone. 
The minerals separate from any given magma in a definite order, which 
is governed by their solubility in that magma, 

Kinds of Minerals. The more important minerals that form the 
igneous rocks are the following: 

Feldspar Group Ferromagnesian Group 

Orthoclase Feldspar, KAlSiA* Mica (Biotite), (H,K) 2 (Mg,Fe) 2 AJ 2 SiA* 

Plagioclase Feldspars Pyroxene, Ca(Mg,Fe)Si 2 O 6 

Hornblende, Ca(Mg } Fe)3Si 4 O 12 
Olivine, (Mg,Fe)*Si0 4 
' Quartz, SiO 2 Magnetite (Iron Ore), Fe 3 4 

Of these minerals, feldspars, quartz, pyroxene, hornblende, and biotite are 
the most important in forming igneous rocks, and consequently the 
student should make careful note of them. For details regarding them 
the Appendix that deals with the minerals mentioned in this work should 
be consulted. It will be seen by examining the chemical formulas given 
in the table above that the minerals are composed of silica and the six 
metallic oxides previously mentioned as occurring in the magmas. 

Furthermore, since it was shown that the chemical composition of 
magmas varies, it is evident that the relative quantities of the minerals 


in the resulting rocks will also vary. Thus, a siliceous magma, in which 
(Na,K) 2 0, A1 2 3 , and Si0 2 are the chief substances, will form a rock that 
consists mostly of feldspars, whereas a basic magma, in which CaO, 
MgO, and FeO are high, will make a rock that contains mostly pyroxene, 
hornblende, and other ferromagnesian minerals, as they are called in 
allusion to the iron and magnesium in them. 

Thus it appears that on account of the diverse compositions of mag- 
mas the igneous rocks that have formed from them vary both in the 
kinds and in the relative amounts of their component minerals. These 
variations in mineral composition largely determine the different varieties 
of igneous rocks, and mineral composition is one of the principal factors 
used in classifying the igneous rocks. 

Classification of Igneous Rocks 

The features by which the igneous rocks may be classified have now 
been explained. We have seen that the igneous rocks vary in texture 
and in composition. Both of these variables will be used as factors in 
building up a classification of the igneous rocks. By employing texture 
as the principal criterion we at once obtain five major classes: I, even- 
granular, in which all the minerals are of about the same size and are 
sufficiently large to be identified by the eye alone or aided by a pocket 
lens; II, porphyritic-granular, in which certain minerals in virtue of 
their large size contrast conspicuously with those which surround them, 
thus forming a porphyry having an even-granular groundmass; III, 
porphyritic-aphanitic, in which the conspicuous crystals the pheno- 
crysts are set in an aphanitic groundmass; IV, aphanitic, in which 
none of the constituents are distinguishable; and V, glassy, in which 
few or none of the constituents have crystallized. These five classes 
are termed massive rocks, and for the sake of completeness Class VI 
is added to take care of the fragmental products of volcanic eruptions. 
This order from Class I to Class V marks in a general way the decreasing 
amount of easily recognizable minerals in rocks: in Class I, all the con- 
stituents are easily recognizable by the unaided eye; in II, the pheno- 
crysts are easily distinguishable, the constituents of the groundmass 
less readily; in III, the phenocrysts alone are distinguishable; and in 
IV, none of the constituents can be recognized. 

Each of the major classes is then subdivided on the basis of composi- 
tion on the kinds of minerals present and the proportions in which 
these minerals occur. It is to these subdivisions that the actual rock 
names are given. For example, an even-granular rock that is composed 
of feldspar, quartz, and a dark mineral, generally biotite, is called granite. 



By applying these principles, the following classification of igneous 
rocks is obtained, as shown in the subjoined table, 


Major Classes 
(based on texture) 

Subdivision of Major Classes 
(based on mineral composition) 

Light-colored minerals, chiefly feldspar, 

Dark minerals 

Dark minerals 


Granular (with 
grams interlock- 

(has quartz) 

(has no quartz) 


(grain size is inter- 
mediate between 
that of gabbro and 



Granular (as 
above) and por- 

(has quartz) 

(has no quartz) 



with aphanitic 

(contains pheno- 
cryats of quartz) 












Volcanic tuff and breccia 

Note: Syenite is briefly mentioned in the text. 

Remarks on the Table. Leaving out of account the glassy rocks, 
which are rare, and the tuff and breccia, which are described on page 255, 
the following remarks may prove of service in understanding the classi- 
fication of igneous rocks shown in the table. 

All rocks in the s^me horizontal column have the same texture. 

All rocks in the same vertical column are essentially of the same 
chemical composition; for example, granite, granite porphyry, and 
rhyolite are alike in chemical composition. In physical appearance, 
however, they differ notably: a granite differs somewhat from a granite 
porphyry, and vastly from a rhyolite. These differences, as already 


pointed out, are mainly the results of the different rates of cooling 
the granite has cooled extremely slowly, whereas the rhyolite has chilled 


The rocks in which the light-colored minerals predominate are light 
in color and light in weight, i.e., they are of low specific gravity. The 
rocks in which the dark (ferromagnesian) minerals predominate are 
dark in color and heavy in weight. The range in specific gravity 
from 2.67 for the average granite to 3 for gabbro is not large, but is 
sufficient after experience to serve as an aid to identification. 

Although in the table each rock has been put in a separate compart- 
ment, in nature no rock variety is sharply bounded from its neighbors 
that are shown in the table. There are for example transitional varieties 
between granite and diorite and between granite and granite porphyry. 
No hard and fast boundaries set off any of the so-called rock species. 
These facts often make it difficult to classify a given rock. It may as 
well be recognized at the outset that the determination of rocks is at 
best a difficult matter, and that to classify accurately with the unaided 
eye the finer-grained and especially the aphanitic rocks, is as a rule im- 
possible. When the accurate identification of a rock becomes a matter 
of high importance, recourse must be made to the microscope. 

Method of Study. The classification that has just been described is 
based on what can be recognized by the eye, aided, perhaps, by a pocket 
lens. It is therefore termed a field classification and 
sometimes megascopic (Greek mega, great), in contrast 
to one based on results obtained microscopically, by the 
study of thin rock slices. Rock slices, or thin sections, 
are made by cementing a chip of rock to a piece of 
glass and grinding it down until the section is one- 
thousandth of an inch thick. In such a thin section, for 
example, the minute mineral grains that make up the 
most fine-grained and blackest of basalts become trans- 
parent, and can be determined under the microscope 

Fig. 196. Thin (pj^ jgg^ j n t ]^ s s t u dy polarized light is used, and a 
section of a roc . g enera i knowledge of minerals, of their crystal charac- 
ters, and optical properties is necessary. It would require too much 
detail to describe further this mode of studying rocks, which combined 
with the examination of them by chemical means has developed into a 
separate geological science, called Petrology, the science of rocks. It 
should be stated, however, that so much additional information has 
been gained by these methods that the precise classification of igneous 
rocks is much more complicated than the simple scheme outlined 


Granite. As may be seen from the scheme of classification, granite 
is composed chiefly of quartz and feldspar (of the variety orthoclase). 
It contains also as a rule a variable amount of flakes of mica, less com- 
monly of hornblende, or both. These component minerals are roughly 
of the same size, and hence granite is said to be even-granular, or equi- 
granular. As the quartz was the last mineral to separate from the 
magma, it is molded around the earlier minerals and occupies the angular 
interspaces between them. This habit of the quartz produces an 
intimately interpenetrating and interlocking arrangement. This inter- 
locking, even-granular texture is so characteristic of granites that it is 
often called for short the granitic texture. It serves to distinguish rocks 
of Class I from all others. The average granite contains 60 per cent of 
feldspar, 30 per cent of quartz, and 10 per cent of dark minerals. There 
are many varieties of granite, based on color, texture, etc. Its common 
occurrence is shown in the fact that there are few states in the Union 
or provinces in Canada that do not contain exposures of granite; and 
its use as a building stone and for various other purposes is well known. 

Granite is the most important intrusive igneous rock, and appears to 
be the main constituent of the foundation of the continental masses. 
These granites are of very ancient origin, of pre-Cambrian age, and con- 
stitute a floor upon which the sedimentary rocks of later age were de- 
posited. Granite of younger age occurs also as stocks and vast batho- 
liths that are intrusive into the younger rocks. In all of these occur- 
rences it is either in the form of normal granite, or in a certain modi- 
fication of it known as granite gneiss. As granite is formed at some 
depth in the crust, it is exposed at the surface only after prolonged 
denudation; hence it" is seen chiefly in those parts of the continents 
bared by erosion that is to say, in mountains or in regions so deeply 
eroded that the roots of the mountains are visible. 

Syenite. This is like granite in texture and composition but differs 
in containing little or no quartz. Several varieties are distinguished, 
based on the character of the feldspar and the accompanying feldspar- 
like mineral. Thus in syenite proper the feldspars are alkalic, that is, 
contain soda and potash, but little or no lime. In another variety, a 
feldspar-like mineral, nephelite (NaAISi0 4 ), is present in addition to the 
feldspars, and the rock is known as nephelite syenite. The syenites are 
not common rocks, nor as a rule do they occur in very large masses com- 
pared with granite. 

Diorite. Diorifce is a granular igneous rock composed of feldspar 
and one or more dark minerals, in which the feldspar is more abundant 
than the dark minerals. The feldspar is mainly plagioclase, but it is 
generally difficult to recognize this fact with the unaided eye. The 


dark minerals may be biotite, hornblende, or pyroxene and they may 
occur either singly or together. 

Gabbro. Gabbro differs from diorite in that the feldspar is subordi- 
nate and the dark minerals predominate. Hornblende, pyroxene, and 
oli vine are the common dark minerals; they occur singly or together; 
biotite, though present in some gabbros, is distinctly uncommon. 
Because of the prevalence of dark minerals, gabbros are dark and of 
high specific gravity. Dolerite is a convenient term for the basic 
rocks that are intermediate in grain size between basalts and gabbros. 

Diorites and gabbros, while abundant as intrusive masses, do not 
commonly occur in extensive batholiths as the granites do. They are 
more common as smaller stocks, sills, dikes, and sometimes forming the 
inner part of thick extrusive masses. 

Peridotite. Peridotite is generally composed of a mixture of ferro- 
magnesian minerals, with olivine peridote, (Mg,Fe) 2 Si04 predom- 
inating. It is not common and usually occurs as minor intrusions, 
dikes, sills, or small stocks. It is very interesting and important, how- 
ever, as being the somrce of ores of chromium, nickel, and platinum, 
and of the diamond. It is generally very dark to black and heavy from 
the large amount of iron-bearing minerals present. The diamonds of 
South Africa occur in volcanic pipes composed of this rock, and they 
have been also found in similar intrusions of it in Arkansas. 

Pyroxenite, as its name implies, is composed wholly of pyroxene, and 
horriblendite consists entirely of hornblende. They form as a rule bodies 
of small size; nevertheless, in places, as at the remarkable platinum 
deposits recently discovered in South Africa, pyroxenite occurs in vast 

Granite Porphyry, Diorite Porphyry, etc. As may be seen by refer- 
ence to the table of classification, there are various kinds of porphyry, 
depending on the coarseness of the groundmass and its composition, and 
on the kinds of minerals that are embedded in it as phenocrysts. Thus 
if the groundmass is as coarse as it is in granite we may have granite 
porphyry, or diorite porphyry. Feldspars are the most common pheno- 
crysts; quartz occurs along with the feldspar phenocrysts, chiefly in 
granite porphyry; and dark to black flakes or prisms of mica, hornblende, 
or pyroxene occur in many porphyries. The porphyries are a very 
common class of rocks, occurring chiefly as minor intrusions: in dikes, 
sills, and laccoliths, and often in necks; they do not occur as batholiths. 
They also compose many extrusive lava flows. Intrusions of porphyry 
in the Rocky Mountain region are very common, and in many places 
are accompanied by valuable deposits of gold, silver, copper, lead, and 
other ores. Examples of this are seen at Leadville and other places in 


Colorado, Montana, Nevada, etc. Porphyries rarely make good build- 
ing stones, as the masses are generally too much divided by joints, but 
in places they serve as excellent road material. A porphyry is shown 
in Fig. 195, B. 

Rhyolite. Rhyolite represents the aphanitic lava form of the magma 
that at depth consolidates as intrusive granite. It contains phenocrysts 
of feldspar, quartz, and biotite, and rarely of hornblende. The number 
of these phenocrysts varies within the widest bounds, so that there is 
every transition between nonporphyritic and highly porphyritic rhyolite. 
When the amount of the phenocrysts exceeds 25 per cent of the volume, 
the rock is by some called a rhyolite porphyry. The colors range from 
white to gray, pink, red, and purple. Rhyolites or andesites with in- 
conspicuous phenocrysts or with sparse or no phenocrysts are termed 

Andesite. Andesites are of many colors, but in general they are 
darker than the rhyolites; dark gray is common. On the one hand they 
are transitional into rhyolites; on the other, into basalt. The average 
or typical andesite occupies the intermediate position. The darker 
andesites are of basaltic appearance, but unlike basalts their freshly 
broken thin edges are translucent when held in bright light. The 
phenocrysts in andesites commonly consist of striated feldspar and one 
or more dark minerals (hornblende, pyroxene, or biotite). Quartz 
phenocrysts are absent (distinction from rhyolite). Andesites that are 
crowded with prominent phenocrysts are by some called andesite 

Andesite and andesite porphyry are enormously abundant among the 
extrusive rocks of the globe. They are the chief products of the vol- 
canoes that form the " circle of fire " surrounding the Pacific Ocean. 
In fact, it was because of their prevalence in the Andes of South America 
that they were given their name. In virtue of their great abundance 
and differences in color, texture, and mineral composition the variety 
of andesitic rocks is legion. 

Basalt. When the color of the lava is very dark gray, dark green, 
brown, or black the rock is basalt, the common effusive equivalent of the 
ferromagnesian magmas. It may be either compact or vesicular. If 
the vesicles have become filled with some mineral, such as calcite, chlo- 
rite, or quartz, the fillings are called amygduks and the rock is termed an 
amygdaloidal basalt. Many basalts are without phenocrysts, but others 
contain numerous conspicuous phenocrysts, consisting of feldspar, oliv- 
ine, or augite, or some combination of these. The phenocrysts are hard 
and have straight, clean-cut boundaries, whereas amygdules are generally 
soft and have irregular, roundish, or elliptical boundaries. The effusive 


occurrence of basalt has been already treated under volcanoes and erup- 
tions. The enormous tracts of land in western America, in India, and 
elsewhere that were flooded by outflows of basalt have there been men- 

Dolerite is the name given to the coarser-grained basalts, in which 
the grains are so well developed that the constituent minerals are nearly 
or quite recognizable. There is no hard and fast line between basalt 
and doterite on the one hand and dolerite and gabbro on the other. 

Felsite. The difficulty and often impossibility of discriminating 
between rhyolites and andesites that are devoid of phenocrysts makes 
it necessary to use an elastic non-committal name. For the light- 
colored rocks of this class, namely those which are white, light to medium 
gray, light-pink to dark red, pale yellow or brown, purple or light-green, 
in short those that are not dark green, dark gray, dark brown, or black, 
the term felsite is often convenient. 

Glassy Rocks. Volcanic glasses occur as thin crusts on the surface 
of lava flows or where a lava flow has been very quickly cooled, and they 
are mostly limited to siliceous magmas. Brilliantly lustrous volcanic 
glass is called obsidian, and the duller and more pitchy variety is called 
pitchstone. Pumice is frothed glass. Obsidian was much used in past 
times by primitive peoples in making weapons, implements, etc. The 
ancient Mexicans were especially skillful in fashioning knives and razors 
from it. Natural glasses, like the obsidian of Yellowstone Park, com- 
monly contain crystallized minerals that occur in spherical forms having 
a radiating or spoke-like structure, known as spherulites. 

Obsidians are commonly dark-colored or even black; and yet many of 
them correspond in chemical composition to rhyolite and granite. 
Hence they appear to contradict the general rule that nearly all rocks 
with this composition are light-colored. However, if a piece of black 
obsidian is chipped to a thin edge it transmits the light and loses much 
of its dark appearance. The deep coloring is the result of uniform dis- 
tribution of a relatively small amount of dark material in the glass. 

Basalt glass is of rare occurrence. Its formation requires extremely 
rapid chilling of basaltic magma. 


1. The Natural History of Igneous Rocks; by Alfred Harker. 385 pages. The 
Macmillan Co., New York, 1909. 

2. Igneous Rocks and Their Origin; by R. A. Daly. 563 pages. McGraw-Hill 
Book Co., New York, 1914. 

3. Rocks and Rock Minerals; by L. V. Pirsson (2nd Ed. by Adolph Knopf). 
426 pages- John Wiley and Sons, Inc., New York, 1926. 



The outer shell of the Earth is not fixed and rigid, but undergoes 
changes that result in movement of some parts with relation to others. 
Evidence is overwhelming that this has occurred repeatedly in the past 
in all places where it is possible to examine the structure of the Earth's 
crust. Movements of the different parts of the outer shell have been 
not only up and down, but also back and forth in directions parallel to 
the Earth's circumference. Evidence of such movements is both direct 
and circumstantial. On many occasions, and even within the present 
century, there have been abrupt, catastrophic shifts through many feet 
or even yards, along local fractures that penetrate deeply into the rocky 
crust. From historic records it is known that gradual movements have 
taken place, with results that are perceptible only after a long time 
interval; and slow changes of this kind are in progress at the present time. 
Back of human history we read the record of crustal movements in 
obvious deformation of the rocks, ranging from broad, gentle bending 
or warping to more localized severe folding or fracturing. 

All movements of the lithosphere, resulting in relative vertical or 
horizontal changes of position and in deformation of rocks, are compre- 
hended under the general term diastrophism. ^Movements that affect 
all or a large part of a continent are termed epeirogenic, from the Greek 
epeiros, a continent!) More localized disturbances related to mountain 
building are designated as orogenic, from the Greek oros, a mountain. 
These terms are useful in discussion; but a systematic treatment of 
deformation may well begin with the more obvious effects, of whatever 
kind, produced either in historic or in late geologic time, and proceed to 
the results of more ancient movements. 


Datum Surface. The fact and the amount of any recent movement 
are determined from the relative positions of features on the Earth. 
Most of these movements are in the vertical direction; and in order to 
determine the extent and the rate of change it is necessary to have a 
convenient horizontal surface to which reference can be made. The 



average level of the sea is the most logical surface for this purpose, as 
the shoreline, at mean tide level, is essentially horizontal throughout its 
whole extent. Any local or differential movement, upward or down- 
ward, of land bordering the sea is clearly evident to one who observes 
the peculiarities of shorelines. 

The idea that the sea surface is undistorted and permanently fixed 
is not strictly correct. Adjacent to high continental borders the water 
is attracted laterally and upward by the land mass, and the water sur- 
face is slightly farther from the center of the Earth in such localities than 
it is along low, flat coasts or adjacent to oceanic islands. Moreover 
the sea level has varied within recent geologic times, first through with- 
drawal of much water from the oceans during the accumulation of 
continental ice sheets, and later through restoration of this water by 
melting of the ice. Further, there are reasons for believing that the 
oceans have increased in size and depth through geologic time by the 
constant addition of magmatic waters; and it is probable that the ocean 
basins have changed appreciably in size many times through upward 
or downward bowing of the floors and by slow filling in of sediment, with 
consequent raising or lowering of sea leveL But such changes are very 
gradual, and their effects in shifting shorelines are essentially uniform 
all over the Earth; whereas many movements of the land are relatively 
rapid, and all such movements vary in amount from one place to 

Elevation. The most striking proofs of uplift of the land consist 
in the locally elevated position of features that we definitely associate 
with the sea or its edge. Thus in various parts of the world outcrops 
of rocks with attached shells or skeletons of dead marine organisms, such 
as barnacles and corals, are found high above sea level. In some lo- 
calities the rocks are pierced by tubes that were drilled and lined by a 
peculiar rock-boring marine animal (Lifhodornus) . Excellent evidence 
of continuous or recurrent elevation is furnished by certain islands of 
the East Indian Archipelago. In shallow waters off the coast of Timor 
corals are building extensive reefs. Similar reefs, strikingly fresh but 
entirely dead, extend from the littoral zone to the higher ground above 
the reach of the highest tides; and still others, showing various degrees 
of weathering, occur at different levels up to several hundred feet above 
sea level. A classic example of changes in land level is found in the 
temple of Jupiter Serapis built by the Romans near the seashore in the 
^einity of Naples. The three columns left standing are bored by litho- 
domi to a height of 20 feet above the floor, and their shells remain in 
some of the holes. From this we infer that after the temple was built 
the land subsided more than 20 feet, carrying the temple into the sea, 


and that later there was uplift of about the same amount. This con- 
clusion is confirmed by historic record. 

Strong testimony is furnished also by the abnormal position of con- 
spicuous features made by erosion and deposition along a coast. In 
parts of California, Chile, Scotland, and numerous other coastal regions 
raised beaches, accompanied by wave-cut and wave-built terraces, 
form nearly level benches of country terminated inland by former sea 
cliffs. Typical wave-formed caves pierce the cliffs (Fig. 197), and old 
stacks, now high and dry, rise abruptly from the terraces. Such an 

Fig. 197. Ancient sea caves in former sea cliff at back of elevated beach, showing 
strand line. Coast of Fifeshire, Scotland. (Geol. Surv. of Scotland.) 

elevated terrace, with its related features, is often spoken of as a raised 
strand line, since it commonly appears as a more or less distinct topo- 
graphic line, or level, approximately parallel to the present shoreline 
and above it. 

Still another kind of evidence, of a direct and positive character, is 
furnished by careful observations made year by year. Thus in countries 
bordering the Baltic Sea an uplift has been under observation for a long 
period, and has been measured by marks placed on the shores. In some 
places the elevation has been as much as 3 feet in a century, but the rate 
is not everywhere uniform and it varies also from time to time. All 
the facts indicate that the Scandinavian peninsula has risen gradually 
for a long period, so that the northern part of Sweden is about 900 
feet higher than it was at the close of the Ice Age. Raised strand lines 



are a noticeable feature in many northern regions (Fig. 198). There is 
similar proof, on a large scale, that within recent geological time the 
west coast of South America has experienced very considerable elevation; 
and probably the uplift is still going on. 

Fig. 198. Elevated strand lines cut in sandstones and limestones. Straits of Belle 
Isle, Labrador. (Schuchert.) 

Depression. Evidence of subsidence below sea level is less striking 
than that of elevation, but not less convincing. It is necessary to use 
care in drawing conclusions, however, because encroachment of the sea 
upon the land is not in itself a proof of subsidence, as it may result merely 
from landward erosion by waves and currents. Submergence of features 
that are definitely characteristic of land surfaces constitutes the best 
proof. Increasing depth of average water level over well-known rocks 
or reefs in harbors gives evidence of slow sinking still in progress. 

Submerged stumps and other marks of former forests are found at 
various places along the Atlantic eoast from Maine southward. Some 
of the stumps remain rooted in the old forest soil, covered by marine 
muds or other deposits (Fig. 199). It is clear that this could occur only 
in situations, such as protected nooks and corners of estuaries, sheltered 
from the waves of the encroaching sea which would otherwise have swept 
away the forest soil. Submerged peat bogs, made under fresh-water 
conditions but now turned into tidal flats, are rather common and tell 
a similar story. All the cumulative evidence goes to show that the 
Atlantic seacoast from Maine southward has gradually sunk within a 
late geologic period. Whether parts of it are still sinking is a matter 
about which geologists have not yet reached agreement. 



Drowned Valleys. Evidence of subsidence of the land is furnished 
by the irregular shorelines produced by the drowning of valleys, with 
production of bays and estuaries. The seaward extension of river 
channels, such as the Hudson, for long distances across the submerged 
continental shelf, demands the same explanation; for manifestly these 
great trenches, now sunk in the sea floor, could not have been cut while 
the continental shelf was covered with water, but only by river or 
glacial action, or both, when it stood at a higher level and was a land 
surface. The submerged channel of the Hudson has been outlined by 
closely spaced soundings on the continental platform south of New 
York harbor. It has the definite form of a valley, extending more than 


Fig. 199. Showing submerged forest, a, old forest soil with stumps sending in it; 
6, marine deposits of silts and sands; c-c, present level of high tide. 

100 miles southward from Sandy Hook to the steep continental slope; 
and its depth below present sea level ranges from about 100 feet at the 
north to more than 2000 feet at the south. 

Subsidence and Deposit of Sediment. In many parts of the world 
thick deposits of sediment are being laid down by rivers in subsiding 
basins adjacent to coasts. Borings into deltas pass through alternating 
marine and fresh-water deposits, or even through sediments entirely 
nomnarine, to a great depth below present sea level. For example, 
wells sunk into the delta of the Po, near Venice, pass through four sep- 
arate layers that contain abundant remains of plants similar to those now 
growing in the marshlands along the Adriatic. One of the layers is 
about 300 feet below sea level At shallower depths some of the sands, 
gravels, and clays contain shells of marine molluscs, and other layers 
yield fresh-water types, such as land snails. In the delta of the Ganges, 
near Calcutta, pieces of wood and bones of land animals are found 
hundreds of feet below sea level. Deep wells in the Great Valley of 
California furnish similar evidence. Facts of this kind show that sub- 
sidence has been going on for a long period, not at an even rate, but as 
an interrupted process, whose variations permitted alternating fresh- 
water and marine deposits to be formed. 

Evidences of Elevation or Depression Inland. Movements of the 
lithosphere involving changes of level are not confined to the sea coasts; 



they occur also in the interior of the continents. For example, in 1811 
large areas of the Mississippi flood plain near New Madrid, Missouri, 
sank far below their former level and now are occupied by lakes. Trees 
that grew on the plain were killed by the flooding, and for a long time 
their dead trunks or tops, projecting above the water, furnished evi- 
dence of the changes in level. 

An excellent illustration of tilting on a large scale is afforded by the 
Great Lakes. To the northeast the land has risen since the retreat of 
the great ice sheet, and as a result the lake basins have been tilted 
southwestward. The old strand lines are several hundred feet above 
the present water level on the north and northeast, and slope toward 
it as they are followed west and south. Since the lakes discharge to the 

Pig. 200. Showing tilting of lake basin. AD, present lake level. BB, raised shore- 
line disappearing under lake at C. Land has been raised on the north (N), and on the 
opposite side the former shoreline, C to E, has been drowned. 

east, the raising of their outlets has caused them to enlarge, expanding 
them to the west and south. As a result, river mouths on the south and 
west sides of some of the lakes (especially Erie and Superior) have been 
drowned. The tilting movement is still in progress and has been ac- 
curately determined. It is at the rate of 6 inches per hundred miles 
per century. Small as this rate seems, in 1600 years it would cause the 
upper Great Lakes to discharge by way of Chicago River into the Missis- 
sippi drainage (Fig. 200) . 

Significance of Recent Changes in Level. It is obvious that changes 
of surface level must be accompanied by bending or fracturing of the 
lithosphere. In fact, surface movements merely record deep-seated 
processes which lead to slight changes in the form of the globe. Com- 
monly the movement is a broad warping that affects wide areas. Locally 
there may be sharper bending, or the rocks may actually break under 
the enormous strain, causing abrupt offsetting of roads or other features 
at the surface. Such breaks, or faults, are common along the Pacific 
coast of North America, in Japan, and in many other parts of the world. 
Displacement along some of these fractures within historic time has 
caused local shifts in level or horizontal changes of position. 


In considering effects of this kind we are impelled to inquire at once 
into their ultimate cause. Discussion of this question will be postponed 
until more of the essential facts and conditions have been presented. 
In general, however, it does not seem strange that a globe so large as 
our Earth, which spins rapidly on its axis and whirls through space, 
which suffers chemical and physical changes in the outer and probably 
also in its deeper parts, should be subject to slow and almost continuous 
deformation. Volcanoes testify to local unstable conditions, which 
result in the melting and moving of rock material. Erosion and sedi- 
mentation through long periods result in the transfer of great loads at 
the surface; and this process undoubtedly sets up enormous strains in 
the lithosphere. These well-known processes would in themselves 
cause some slow deformation of the crust; and probably more profound 
changes in progress in the deep, unknown parts of the Earth are re- 
sponsible for still greater deforming forces. 

Regardless of ultimate causes, it is desirable that we get from the 
actual facts a picture of the Earth as a changing, dynamic thing. " Ter- 
ra firma " is not literally fixed and inert. Parts of the Earth's crust 
are changing in form and position today, and all parts have moved 
at some time during geologic history. 


It is neither desirable nor possible to distinguish sharply between 
recent and older deformation. Certain features at the surface bear 
unmistakable witness to very recent crustal disturbance, although there 
may be no direct record of the event in human history. On the other 
hand we see in the rocks many fractures and folds that date from very 
early periods in Earth history, as is shown clearly by geologic relation- 
ships. Between these two extremes are found indications of disturbance 
in every geologic epoch, showing that movements of the same or similar 
kinds have been continuous or recurrent throughout recorded geologic 
time. In general the record of last movement is found in forms and 
features on the Earth's surface, such as the elevated beaches, drowned 
valleys, and tilted lake basins described above. All surface forms are 
ephemeral, because of erosion; and as they disappear or become frag- 
mentary, the most reliable guide to former crustal movements is found 
in the structure of the underlying rocks. 

Study of the framework or structure of the lithosphere is called struc- 
tural geology. Such a study may be entirely geometrical, with the ob- 
ject of furnishing an exact description of the crust, including all breaks 
and bends in the rock units. Knowledge of this kind has great economic 


value, as it Is essential in locating and following beds of coal or mineral 
veins. However, the structural features of the Earth have also a broader 
interest and value, as they furnish the clue to important events in the 
history of a region. 

The principal structural features that indicate deformation of the 
lithosphere are broad bends or warps 7 folds of various sizes and degrees 
of intensity, and fractures. 


Certain topographic features that result from gentle warping of the 
crust develop slowly and persist for a very long time. For example, 
when a peneplain is domed up widely the streams become deeply en- 
trenched and their old meander patterns are preserved in the deepened 
valleys. As the deepest entrenching results in the area of greatest up- 
lift, the middle portion and the edges of such a great dome can be recog- 
nized by study of the entrenched valleys. Furthermore remnants of 
the uplifted peneplain surface will persist for a long time between the 
valleys. By study of the valleys and the remains of old surfaces it is 
determined that the Appalachian region is an irregularly domed pene- 
plain undergoing dissection. It is obvious that evidence of this kind, 
used for recognizing warping movements that are quite old, is similar 
to that by which recent uplift is determined. 

A more permanent record of gentle warping is found in the bending 
or tilting of stratified rock formations that originally were nearly or quite 
horizontal. In southeastern Texas marine limestones and shales de- 
posited in Cretaceous time now lie slightly above sea level, with a gentle 
inclination toward the Gulf. Formations of the same age, containing, 
fossils of identical marine forms, are exposed in eastern Colorado, 5000 
feet and more above sea level. When these beds were laid down a 
continuous seaway reached northwestward from the Gulf of Mexico 
across the United States. In later time this part of the continent was 
elevated by a broad warping movement, so that the formations in Colo- 
rado were lifted a mile higher than the corresponding beds in Texas, 
700 miles away. In the Colorado Plateau of Utah and Arizona a thick 
blanket of marine strata lies thousands of feet above sea level, and is 
dissected by deep canyons. In a general way these strata are nearly 
horizontal; but if any one layer is followed and mapped in detail it is 
found to bow gently upward into irregular domes and bend downward 
into shallow basins. It is clear that this wide area of sedimentary beds, 
many hundreds of miles across, was not lifted up with absolute uniform- 
ity, but was bent somewhat, so that each layer now resembles a wide 
board that has been warped by exposure to the weather. The original 



surface of the uplifted mass has been entirely destroyed by erosion; 
but a record of the distortion is preserved in the form of each layer of 
stratified rock beneath that former surface. 

The value of widespread marine sediments as a means of detecting 
such gentle warping of the crust is obvious. In areas occupied entirely 
by granites and similar massive rocks, a broad warp is recorded only 
in the surface forms; and after these forms have been obliterated by 
erosion there is nothing to suggest the ancient movement. 


Anticline and Syncline. In many places the stratified rocks have 
been buckled into plications or folds. Some of these are on a small 
scale and can be seen directly; but commonly the folding is on such a 
great scale, and exposures of the rocks are so discontinuous, that it is 
necessary to study and piece together the structure of certain distin- 
guishable layers over many miles before the form of the folds becomes 
clear. The nature and scale of the folds in mountain regions can be 
appreciated best by study of maps 
and cross sections that have been pre- 
pared by geologists in the field. 

Two terms are used constantly in 
descriptions and discussions of folds. 
Like regular swells on the ocean, 
rock folds ordinarily occur in a 

series, with alternating crests and troughs. The crests of the folds 
that is, the upfolds are anticlines; the troughs, or downfolds, are 
synclines (Fig. 201). Even if the original surface crests should be carried 
away by erosion, and the whole reduced to a level plain, we should still 
call the upfolded portions below the surface anticlines, the downfolded 

Fig. 201. To illustrate anticlines, A, 
and synclines, S. 

Fig. 202. Anticlines, A, and synclines, S. Folds have no relation to present 

surface forms. 

portions synclines, and in imagination reconstruct the missing parts 
(Fig. 202). Thus it should be clear that anticlines and synclines are 
not a matter of surface topography, but of structure (Figs. 203 and 204). 
Commonly the original configuration of the surface is even reversed by 
erosion, so that valleys now occupy the positions of the ancient crests, 
and ridges or mountains are in the places of the troughs; but the original 
structural terms still apply (Fig. 202). 



Fig. 203. An anticline broken at the top. In the foreground the outcrops of the 
eroded strata are seen dipping outward, from which the anticline structure could be in- 
ferred if the arch did not exist. Pembroke, Wales. (Geol. Surv. of England and Wales.) 

', ' ," '//',#.*:' "v " '!: * ; -" v '4^',2v'5 -';' * *' \, 

. l -. ; 

Fig. 204, A syncline, near Hancock, Md. (U. S. Geol. Surv.) 


Outcrop. Only the ideal relation of simple, upright, regular folds 
has been considered above. A series of folds approximating this form 
is by no means uncommon in nature; but usually the folding is much 
more complicated. Moreover some important facts about folds cannot 
be represented in ordinary cross sections. The varied kinds of deforma- 
tion which the rocks have suffered in any region condition the geo- 
logic structure of that region; and it is a matter of the highest importance 
that the geologic structure of every country should be known so far as 
possible and represented accurately on maps. If the surface of the 
Earth were everywhere naked bedrock, this would be relatively an easy 
matter; but since the rocks have been greatly eroded, and are largely 
covered with earth and vegetation, or with water, snow, and ice, the 
natural difficulties of the task are enormous. The structure in a region 
is determined by a careful study and comparison of the outcrops, by 
which term is meant the actual exposures of bedrock at the surface. 
If the ground were perfectly level and 
the strata horizontal, the outcrop 
would be the flat surface of the upper- 
most rock stratum, and we should 
learn little from it; but on slopes and 
cliffs bordering stream valleys, we 
may inspect the outcropping edges of Fig. 205. Section and outcrop of 

. , TJ? , ! . i f . i horizontal strata along valley. 

many strata. If the sides of the 

valley were trenched by ravines, the line of outcrop would not be 
straight, but sinuous, retreating from the valley into the ravines, and 
advancing on the spurs (Fig. 205). 

If the strata have been inclined by folding and eroded their edges may 
be exposed even on a nearly flat land surface. Commonly the edges 
of the harder, more resistant beds project to form the more prominent 
outcrops. In mountain regions, soil and other concealing debris 
usually decrease in amount with increasing height; and exposures of 
rock grow in prominence correspondingly, until the upper rocky ridges 
and peaks may each be a vast outcrop. Because of the excellent ex- 
posures, and the great depth of the section visible in canyon walls and 
on the cliffy slopes, mountains furnish the most favorable opportunities 
for determining geologic structure. 

Dip and Strike. These terms, used constantly in describing the 
attitude of inclined strata, may be defined as follows: Dip is the angle 
of inclination of the plane of bedding from a horizontal plane. Strike is 
the direction of the line of intersection of the plane of bedding with a hori- 
zontal plane. Reference to a diagram will make the definitions clear. 
In Fig. 206 any horizontal line, as AB, drawn on the surface of the in- 



clined stratum represents the strike. Lines with any other position 
on this surface, as EF or GH, are inclined; but CD, at right angles to 

Fig. 206. To illustrate strike and dip of tilted strata. The line of intersection of the 
horizontal water surface with a bedding plane gives the strike; or any other horizontal 
line on the bedding, as AB, also represents the strike. The lines EF and GH have an 
inclination, but the line CD, which is at right angles to AB, shows the maximum or true 
dip of the layer. 

the strike line, represents the maximum inclination of the surface, or 
the dip. The practical use of strike and dip are better illustrated in 
Fig. 207. A dark stratum, AG, is exposed on a perfectly flat surface. 

Fig. 207. To explain strike of strata, and direction and amount of dip. The stratum, 
A-G, is exposed on a horizontal surface. The direction of the strike, with relation to the 
compass points NESW, is shown by the angle NOK, which is east of north. The stratum 
dips in the direction OD t which is east of south by the amount of the angle SOD. The 
angle APH, measured from the horizontal plane, is the angle of dip. 

The direction of its outcrop, PG, is the strike. If this direction, meas- 
ured by the angle NSG ( = NOK), is 30 degrees east of north, the stratum 
is said to strike N. 30 E. Obviously a layer with this strike might dip 



either toward the northwest or toward the southeast. The stratum in 
the diagram inclines southeastward, making an angle APH with the 
horizontal. If this angle measures 40 degrees, the stratum is said to dip 
40 southeastward; or more precisely, as the direction of dip is at right 
angles to the strike, and the latter is N. 30 E., we say the dip is 40 S. 
60 E. In the diagram, the direction of the dip (S. 60 E.) is shown by 
the angle SOD. The meaning of the statement is clear if it is kept in 
mind that the first angle 40 is the amount of the dip, measured 
downward from the horizontal (angle APH, Fig. 207); whereas the 
second angle 60 gives the direction in which the stratum dips, 
with relation to the north-south line. 

The direction of strike is taken with an ordinary compass, and the 
direction of dip is calculated. The amount of dip is taken with a clino- 
meter, which is essentially a pendulum swinging over a graduated arc 
(Fig. 208). For geologic purposes the compass and clinometer are usu- 
ally combined in one instrument. 

Dip and strike are represented xm geologic maps by a conventional 
sign ~, in which the direction of the cross bar, as placed on the map, 
indicates the direction of strike, and the arrow points in the direction of 

Fig. 208. To explain measurement of dip angle with the clinometer. The pendulum 
of the instrument swings freely on an axis, and therefore is always vertical when the box 
is on edge. When the edge of the box rests on a bedding plane in the direction of dip, the 
angle of dip is read directly on the graduated arc. 

dip (Fig. 209). The length of the arrow is also sometimes used to show 
in a general way the amount of dip; thus |- indicates a low angle of 
inclination; |- a steep dip. Ordinarily the actual amount in degrees 
is written in; e.g., H> 30. 

Pitching Folds. A series of dips and strikes arranged on a map as 
shown in Fig. 209 indicates a set of parallel folds in the underlying rock. 
It is quite evident that folds of this kind could not run in the direction 



of strike indefinitely, or around the world; they must end somewhere. 
The end of a oyncline, as seen on a map and in section, has the appear 

Fig. 209. Diagram of a land surface underlain by folded strata. In making a geologic 
map of this area, the strike and dip of each outcrop is noted in the proper position, by use 
of the conventional symbol ~. The symbols in the figure indicate the positions of an 
anticline and two synclines, which are shown in section at the end of the block. 

ance shown in Fig. 210, A. A stratum is warped into a form like the end 
of a boat. An anticline, near its termination, resembles a boat over- 
turned (Fig. 210, B). In either of these structural features the outcrop 



Fig. 210. A, ending of a syncline, as seen on a flat surface. B, near the ending of an 
anticline. For simplicity, only a single stratum is shown in each case. Consider that 
each block is about half a .mile wide. (H. H. Robinson.) 

of a stratum, on a nearly flat surface of erosion, turns in the shape of a 
horseshoe; but there are two general ways for distinguishing one form 
of fold from the other. In the ordinary syncline, dips are consistently 
toward the inside of the horseshoe (Fig. 211, A), and the younger strata 



lie inside. In an ordinary anticline, dips are outward at every point, 
and the older strata lie inside the horseshoe (Fig. 211, B). 

The median line of a fold, along the top of an anticline or the bottom 
of a syncline, is the axis. This line extends along a bedding surface, or 
along this surface restored if it has been partly eroded (Fig. 210). The 
axis emerges from the ground toward the end of a syncline, plunges 
into the ground toward the end of an anticline. The angle between the 
axis and the horizontal is called the pitch of the fold. It is evident that 
the pitch is merely a special case of dip measured along the axis. In a 

Fig. 211. Illustrating the use of dip and strike symbols on a map to show a plunging 
syncMne (A) and a plunging anticline (B). 

pitching or plunging fold, the sides or limbs, as exposed on a nearly level 
surface, cannot be parallel, but necessarily converge or diverge (Figs. 
210-212). If outcrops of strata in the two limbs run along parallel to 
each other for a long distance, the axis of the fold must be essentially 
horizontal. In the Appalachian region, hard sandstones or other re- 
sistant formations in the limbs of eroded folds make some of the high 
mountain ridges. Commonly two such ridges, on opposite sides of a 
great fold, maintain a straight parallel course for many miles; but 
eventually they converge and unite at the end of the fold. 

Inclined, Asymmetric, and Broken Folds. Thus far we have con- 
sidered only simple, regular, upright folds. If a plane is imagined to 
pass through the center of a fold and its axis, as in Fig. 213, like the 
extended keel of a boat, we may call this the axial plane of the fold. 
In a regular or symmetric fold, this plane is one of symmetry; that is, 
the parts to left and right of it are symmetrically disposed, or each point 
on the left of the plane has its corresponding point at an equal distance 
on the right of it. If the fold is upright the plane is vertical (Fig. 213). 
However, some folds are not upright but have been pushed over until 
the axial planes are inclined. A fold of this kind is said to be over- 



turned (Fig. 214). Such overturning may, indeed, go so far that the 
axial plane is nearly, or actually, horizontal; and the fold is .then termed 
recumbent. Synclines may have a similar attitude. 

Fig. 212. On a plain of marine erosion the outcropping edges of the strata are seen 
at the ending of a syncline, as shown by the curving strike and inward dip. Near North 
Berwick, Scotland. (Geol. Surv. of Scotland.) 

It is also common to find that folds are asymmetric (without sym- 
metry) ; that is, they are not similar to right and left of the axial plane, 

Fig. 213. Upright symmetrical 
fold; axial plane vertical. 

Fig. 214. Inclined symmetrical 
fold ; axial -plane inclined. 

which is not, therefore, one of symmetry, as in a regular fold (Fig. 215). 
Such asymmetric folds may be upright, overturned, or recumbent. 

Finally, folds may be so sharply flexed (creased) that they may break, 
especially at the apex; and on breaking, the parts are likely to be dis- 



placed with respect to one another, or faulted. Faulting, however, is 
so important a phenomenon that it deserves especial consideration in a 
later place. 

Other Features of Folds. If folds are so sharply flexed that the 
limbs are nearly or quite parallel, they are said to be dosed; in this 
condition the horizontal distance across the strata> or the width of the 
fold, cannot be farther reduced without squeezing or mashing of the 

Fig. 215. An asymmetric fold. 

Fig. 216. A, closed fold; B, part of an 
open fold. 

beds (A, Fig. 216). If the limbs make a large angle with each other 
(as in B, Fig. 216) the fold is open and the strata may be further folded 
without mashing. 

In isoclinal (equal inclination) folds the strata are compressed until, 
on both sides of a fold, and perhaps throughout a series, they are parallel 

Fig. 217. Outcrop of strata show in cross section as in a; they might be one series 
with inclined dip, or possibly are in a closed isoclinal fold. Assuming that the strata are 
arranged symmetrically on opposite sides of a middle line, the structure is an anticline, 
as indicated in &, if the strata are progressively older toward the middle ; or it is a syneline, 
as shown in c, if the strata are progressively younger toward the middle. 

and have the same dip (Fig. 217, b and c). When such folds are cut away 
by erosion as in a, some skill is required for correct interpretation of the 
structure. The term homodine is sometimes used to describe a series 
of bedded rocks all dipping in the same direction. The strata may be 



folded isoclinally, or simply tilted uniformly. A special kind of homo- 
cline is the monocline, in which the strata are bent in one direction only 
(Fig. 218). A true monocline is a one-limb flexure, on either side of which 
the strata are horizontal or have uniform gentle slopes. 

Folds have been treated thus far as simple structures with true axial 
planes; but many folds are warped or bent, so that they do not have 
axial planes in the true sense of the word, as the surface bisecting each 

Fig. 218. A monoclinal fold. 

fold is not plane but curved into a sinuous form. Folds may also branch 
into compound, complicated structures. 

Geosynclines and Geanticlines. Long belts within a continent or on 
the ocean floor have been warped down to form geosynclineSj whose 
dimensions are measured in hundreds of miles. Correspondingly great 
upwarps are called geanticlines. The prefix in each case (from the Greek 

Fig. 219. Illustration of terms used in compound folding. The general uplifted 
masses of folds A A are called anticlinoria, while the, downwarped mass of folds S is termed 
a synclinorium. The general average warping effect of the folding is indicated by the 

word Geos f meaning Earth) emphasizes the scale of these features. In 
contrast with ordinary folds, the flexures responsible for geosynclines 
and geanticlines are very gentle. The Baltic Sea may be a modern 
geosyncline. In the region about Cincinnati, Ohio, the Paleozoic 
strata are bowed gently into a geanticlinal arch 250 miles wide. 
The geosynclines of the past, as well as those of the present, have been 


the great basins for the accumulation of sediments, like those exposed in 
the Appalachians and the Alps. When later the accumulated beds are 
compressed into folds, the whole series may form a compound uplifted 
mass, which erosion carves into mountains. Such a mass of strata, laid 
down in a geosyncline and crushed into folds, has been termed by 
Dana a syndinorium (from syncline and oros, Greek for mountain). 
The term thus introduced by Dana has, however, been diverted from its 
original meaning, and applied to a general syncline compounded of minor 
folds and contrasted with antidinorium (Fig. 219). It has thus become 
a term of structure, and the related idea of mountain making, which 
the name expresses, has been relegated to a subordinate position, or 
entirely left out. 


In the outer shell of the Earth the rocks are traversed in all directions 
by fractures, varying from minute crevices to important fissures. We 
have considered the importance of fractures in the weathering of rocks 
and formation of soil; in the holding and in the circulation of ground 
water; and we shall discuss them again in connection with mineral veins. 
They are, indeed, of great geologic importance, because of the processes 
that give rise to them 'and the results achieved by their aid. 

A fracture on which there has been no appreciable displacement is 
called a joint. Ordinarily joints are closed so tightly that little or no 
space is visible between the walls. If the walls are distinctly separated, 
the term fissure is preferable to joint. Some fissures are open, and others 
have been filled with mineral matter deposited by circulating water. 
If there has been relative displacement of the walls in a direction parallel 
to the fracture, so that corresponding points on the two sides are dis- 
tinctly offset, the fracture is known as a fault. Faults are very impor- 
tant geologic features. 

Joints in Stratified Rocks. Field examination shows that joints 
are common, but that they are much more numerous in some places than 
in others. Where they are abundant, commonly they are arranged in 
more or less definite sets; that is, the divisional planes running through 
the rock fall into groups according to direction. In many places there 
are two prominent sets of joints, approximately at right angles to each 
other and each set nearly vertical. Such a combination of two or more 
intersecting sets constitutes a joint system. Combined with natural 
divisional bedding planes, a well-defined system of joints divides strati- 
fied rocks into series of closely fitted blocks. The finer the grain of the 
rock, as a rule, the more perfect the jointing and the more definite the 



resulting blocks. Thus, in shale beds and in limestones, the jointing 
may be very perfect, as illustrated in Fig. 220. 

ijSuch jointing may result from various forces; for example, from the 
tension produced in the beds of sediments by contraction when they are 
elevated from the sea-bottom to form land masses, and undergo a drying- 
out process. A more common probable cause of regular fracturing is 
the warping and twisting suffered by the strata during crustal move- 

Fig. 220. Illustrating joints in limestone beds. The horizontal lines are bedding 
planes. There are two sets of joints, nearly vertical and at right angles to each other. 
Dnimmond Island, Mich. (U. S. Geol. Surv.) 

ments.*) The exact cause of most joints in stratified rocks is not surely 
known; but they have sometimes been classified as tensional or com- 
pressional, according to the supposed nature of the force producing them. 
In regions where the strata have been definitely folded, as in many 
mountain zones, unquestionably the compressive force that plicated the 
beds produced many joints. More locally, and especially at the crests of 
anticlines, joints may have been made by stretching or tension. 

Strike joints are parallel to the strike of the beds, or nearly so; dip 
joints are essentially in line with the dip, and therefore at right angles 
to strike joints. Oblique joints have intermediate directions. Certain 
joints extend for long distances across a thick series of beds, and are 
known as master joints. They are contrasted with minor fractures, 
which may be limited to a single stratum. 

Joints in Igneous Rocks. -{-A common type of jointing in igneous 
rocks is due to the contraction resulting from the cooling of the original 


magma./ This occurs during and just after solidification from the liquid 
state. It may manifest itself in one of several ways, depending on the 
rate of cooling, the size and shape of the igneous body, and other factors. 
Thus intrusive masses of granite and similar rocks are characteristically 
cut by joint planes that divide them into large blocks or prisms. Finer 
grained masses in sheets, laccoliths, and dikes may be divided into small 
angular fragments by closely spaced joints. In some laccoliths and 
similar dome-shaped intrusions there is a shelly jointing on a large scale, 

Fig. 221. "Devil's Post-pile"; columnar jointing in lava. Head of San Joaquin 
River, Calif. (XI. S. Geol. Surv.) 

parallel to the domed surface. This appears to have been caused by 
nearly uniform cooling of the mass from the periphery, with resulting 
separation into sheets. 

The most striking kind of contraction jointing in an igneous rock 
results in development of columnar structure. This result is produced, 
in general, when two dimensions of the mass are great and the third is 
small, as in a dike, a sill, or a lava flow. The rock-body may then be 
composed of a series of closely fitted prisms, which are subdivided *by 
inconspicuous cross joints. The prisms have a variable number of 
sides, but commonly they tend to be hexagonal, and some of them have 
remarkable regularity of form (Fig. 221). They may range from several 
inches to a number of feet in diameter, and up to 200 feet or even more 
in length. The Giant's Causeway on the north coast of Ireland is one 
of the most celebrated examples of this columnar structure. The col- 


umns form at right angles to the chief cooling surface, and consequently 
in a level intruded sheet or flow of lava they stand vertically, whereas 
in a vertical dike they tend to be horizontal. Thus some dikes, exposed 
as walls by erosion, resemble regularly piled cordwood. In other igneous 
bodies the position and form of the columns depend on the directions 
taken by the periphery of each individual mass. Columns in volcanic 
necks may be vertical, horizontal, or curved; and in some masses they 
may be arranged radially, like a great fan. 

The reason for columnar structure appears to be that in a cooling 
mass, centers of contraction tend to occur on the cooling surface at 
equally spaced intervals. From each center three cracks form and 
radiate outward at angles of 120. Intersection of these cracks pro- 
duces a regular hexagonal pattern, and their penetration inward makes 
the columns. But nature is complex, and the ideal pattern is commonly 
modified by the occurrence of five- and four-sided figures of varying 
dimensions. By contracting lengthwise the individual columns may 
break into sections. The same principles of contraction result in the 
polygonal shapes commonly seen on mud flats that have cracked from 
drying (Fig. 155). 

In addition to the joints caused by cooling, later fractures caused by 
crustal movements may affect igneous rocks; but wherever a prominent 
columnar structure or other well defined fracture system is original in 
the rock mass, later stresses are more likely to be relieved along these 
existing breaks than to form additional fractures. 

Jointing in Metamorphic Rocks. As a rule the metamorphic rocks 
are much jointed. This might be expected, because 'of the extensive 
deformation to which such rocks have been subjected. The character 
of the jointing varies considerably with the nature of the rock. Many 
of the massive gneisses have joint systems like those characteristic of 
granite; whereas the fissile and schistose rocks, such as slates, have 
joints more like those found in sedimentary rocks. (See Chapter XIV.) 

Practical Importance of Joints. Joints are a matter of great im- 
portance in all quarrying, tunneling, and mining operations where rock 
work enters as an important factor, since the jointing obviously facili- 
tates progress. Without them, every rock fragment would have to be 
broken or blasted loose from bedrock. However, joints may ajiso be a 
serious inconvenience, especially if large blocks of quarried stone are 
desired. Perfect monoliths 50 or 100 feet in length can be obtained from 
comparatively few localities. 

General Features of Faults. Displacement of rock masses along a 
fracture may occur at the time of the break, or at some later time. 
Thus a joint might eventually become a fault. Faults are common 



features in rocks of all kinds. They are most evident in stratified forma- 
tions, as the offsetting of layers makes the break conspicuous and di- 
rectly measurable. However, massive igneous rocks may be faulted 
as well; and as mineral veins or other features of economic value may be 
displaced by such fractures, it is important from a practical as well as 
a scientific standpoint that the nature of faults be well understood. 

The surface of fracture along which movement and dislocation has 
occurred is often spoken of as the fault plane. Although a limited part 
of it may be nearly plane, it is rarely flat for any considerable distance, 

Fig. 222. Part of an old fault surface, with slickensides, uncovered by erosion. Note 
the lined and fluted character, indicating that the movement was directly down the dip 
of the surface. Spotted Range, Nevada. (Longwell.) 

but more or less curved, broken, and offset. Therefore it is better, and 
causes less misapprehension, to term it the fault surface. Moreover, the 
movement in faulting may occur, not upon one surface, but upon a 
number of more or less closely adjacent breaks, producing a fault zone, 
in which the various offsets make in the aggregate the total displacement. 
Such a distribution is sometimes called step faulting. The masses of 
rock involved in fault movements are generally of such size and weight, 
and so compressed together, that the motion of one fault face on the 
other takes place under tremendous pressure. As a result of the friction, 
the rock faces are smoothed and striated, and not uncommonly receive 
a high polish. Such polished and grooved surfaces are known as dich- 
ensides (Fig. 222). The line of intersection of the fault with the plane of 
the horizon is called tine. strike, or trend, of the fawlt, just as we speak 
of the strike of upturned strata. The surface of faulting is rarely ex- 



actly vertical; commonly it is inclined, and in some important faults it 
approaches horizontality. The angle between the fault surface and the 
horizontal plane is the dip. In an inclined fault the side that overhangs 
is known as the hanging wall, the other as the foot watt (Fig. 223). If 
one were to descend along a fault, as in an inclined shaft of a mine, the 
appropriateness of these old mining terms would be evident. 

Generally the fracture is closed tightly; but parts of it may have been 
open at one time, and have been filled with mineral matter deposited 
from solution. Along many faults the grinding of the walls upon one 

Fig. 223. To explain fault relations and terms. The strata have been displaced by 
a fault, and a vein of mineral (black) has formed along the fracture. Mining operations 
have removed the mineral to a considerable depth, exposing the hanging wall and foot 
wall of the fault. 

another has produced a zone of broken and crushed rock known as 
fault breccia. Commonly there is a thin seam of clay-like material, 
known as gouge, directly along the fault. In the displacement of strati- 
fied rocks the friction usually causes bending of the layers near the fault 
surface. This feature, referred to as drag, may be a useful aid in de- 
termining the relative direction of motion on the two sides of the fault 
(Fig. 224). 

The features explained so far have to do chiefly with faults as seen 
below the surface of the ground. Ordinarily a fault breaks the surface 
as well as the rocks beneath; and if one side of the fault is elevated with 
relation to the other, the result is an abrupt cliff, or fault scarp. With 
the passage of time the original scarp is modified or even removed by 

Motion on the Faplt. If we assume that one side of a fault stands 
fast, motion on the other side may be vertical, horizontal, or oblique. 



Thus in Fig. 225 the lettered plane may represent one fault face, which 
for convenience is considered to remain at rest. It is exposed by re- 
moval of the block which is assumed to have moved. If we suppose that 

Fig. 224. Fault in shale; the drag of the beds shows that the left side has gone down, 
the right up. Little River Gap, Tenn. (U. S. Geol. Surv.) 

some particle at A, for example a crystal, was cleft by the fault, then one 
part A remained in its original place and the other part, embedded in the 
opposite face, was carried in some direction by the faulting. Suppose 
that the moving portion of the crystal scratched a groove on the station- 
ary fault face. Depending on the direction of 
movement, this scratch or striation might be one 
of the lines AH, AG, AC, or AD. Most com- 
monly it would be in some oblique direction, as 

Normal and Reverse Faults. If faulting 
takes place by movement upward or downward, 
two different kinds of structure may result. In 

A, Fig. 226, the hanging waU has apparently fflDie motion m laumn? 
slipped down with reference to the foot wall; a The hanging-wall block is 

r- -i. J.T-* i * j i ^ ,v,/vww,^7 -Foiil-f omitted, to show entire fault 

fault of this kind is known as a normal fault. surface / 
In B of this figure, the hanging wall has appar- 
ently been crowded up over the foot wall; a fault of this kind is called 
a reverse fault. In the normal movement a particular layer V-V has 



been lengthened apparently by an amount corresponding to the gap 
C-E] in the reverse movement it has been shortened by an equivalent 

It should be realized that the movement of one wall or the other to 
produce a normal or a reverse fault is purely relative. Thus a normal 
fault may result from keeping the hanging wall stationary and moving 
the foot-wall block upward; or both blocks m$y move, in opposite. 

Fig. 226. A. Simple normal fault. B. Simple reverse fault. 

directions. In nature we see omy the effect of faulting, and the actual 
character of the movement must be inferred. It is even possible that 
some faults described as normal or reverse were produced by horizontal 
slipping. The tilted layer BB, in Fig. 227, A, is offset by pushing the 
front block to the right. In a vertical section, such as might be exposed 
in the side of a canyon (Fig. 227, B), it appears that the hanging wall has 
moved down with relation to the foot wall; in other words, the fault 
would be described as normal. If the front block had been displaced 
to the left, a vertical section would show a reverse fault. 

Components of Faulting* It is often necessary for purposes of 
description or measurement to resolve a fault into component parts. 
On the diagram, Fig. 228, AGEF represents the horizontal plane and A-F 
is the strike of the fault". Let us suppose that the motion has been such 



that a particle near A has been carried to the position B] then the line 
A B joining these two positions represents the displacement, or slip; 
no matter what actual path the particle may have followed, AB is the 

Fig. 227. To illustrate normal faulting, as seen in a vertical plane PP, caused by simple 
horizontal shoving on the fault surface FF. The particular stratum B in PP (right-hand 
figure) appears to have slipped down. Modified from Ransome. 

resultant, and its length the measure of the slip. The line AB, how- 
ever, in order that it may be fixed and determined, must be referred to 
known axes or planes. The line AC gives the amount of motion in the 
horizontal direction at right angles to the strike FA, and this is known as 

Fig. 228. To illustrate and define the components of a fault. 

the heave of the fault; the line CH is the amount of vertical motion and 
is called the throw of the fault; HB(=AM or CD), the amount of motion 
in the horizontal direction along the strike, is termed the strike-slip of 



the fault. Thus there are the three right-angled axes AC, DC, and 
HC, meeting in the common point C, and these may be termed the com- 
ponent axes of faulting. The directions and intercepts on these axes 
being known, the displacement can be calculated, and the problem of the 
fault solved. 

The heave and throw of faults are the components commonly recog- 
nized, because the dislocation is most easily seen in a vertical section at 
right angles to the strike of the fault, and in this section the particle 
A has apparently moved from A to H. The strike-slip is difficult to 
estimate in most faults and often it cannot be determined at all. 




Fig. 229. Bhistrating strike faulting in stratified rocks. A, before f aid ting; B, after 
faulting, fault scarp still uneroded; C? surface levelled by erosion. 

It is clear that a fault might take place without strike-slip, the move- 
ment being wholly down the dip of the fault surface. If the fault were 
exactly vertical, obviously there would be no heave; there might be 
strike-slip, but this also might be wanting and the fault would have 
throw only. Whatever the dip of the fault, conceivably the movement 
might consist of strike-slip only, without either throw or heave. 

Faults in Stratified Rocks. Although faults occur in all kinds and 
combinations of rocks, they show to best advantage in stratified beds, 


Kg. 230. Illustrating dip faulting: A, before faulting; B, after faulting, fault scarp 
uneroded; C, surface levelled by erosion, showing offsets of strata. 

on account of the strongly marked stratification which they disarrange. 
Certain terms are used to define faults in relation to the structure of 
the sedimentary beds. Thus in a strike faulty the strike of the fault and 
that of the strata are parallel, or nearly so, as illustrated in Fig. 229; 
dip faults cut directly across the strike of the strata, or nearly so, as 
shown in Fig, 230; oblique faults cut diagonally across the strike of the 



strata. The figures show only normal faults; but they may also, of 
course, be reverse. The figures also indicate no real strike-slip, but in 
Fig. 230, C there is offsetting of beds, with a false suggestion of strike 
movement. Such abrupt offsetting of tilted strata is one of the surest 


<( ^j 

Fig. 231. Repetition of formations by a normal strike fault, a and a' are parts of 
the same limestone member; 6 and , parts of the same shale member. The fault runs 
through F-F', and dips to the right. M,uch of the hanging-wall block has been removed 
by erosion. Throw of the fault, 300 feet. Spotted Range, Nevada. (LongwelL) 

indications of dip or oblique faulting. Strike faults are more difficult 
to perceive and may easily be overlooked; they may cause deception 
as to the thickness of strata by producing repetitions (Fig. 229, C and 
Fig. 231). Thus, in traversing strata the repetition of a certain set 
should lead to suspicion of strike faulting. 
On the other hand, strike faults may conceal 
strata after erosion has occurred. Thus in 
Fig. 232, which represents a reverse fault 
with later erosion, there is no outcrop of the 
stratum A at the surface. 

The movement of one side of a fault on 
the other side may be attended by rotary Fig> 232 . __ illustrating con- 

Or pivotal motion, as illustrated in Fig. 233. cealment of strata by strike 
.;,.,,. , . , , faulting and subsequent removal 

A fault of this nature is known as a rotary of the scarp by erosion, 
fault; or, if the displacement dies out grad- 
ually up to a definite point, it is a hinge fault. After erosion has 
levelled the surface, a fault of this kind is indicated by a pronounced 
difference in the strike and dip of strata on opposite sides of the break. 
Some faults pass gradually into monoclinal folds (Fig. 234). 



The Magnitude of Faulting. The scale on which faulting has taken 
place varies within the widest bounds. The displacement may be but 
a fraction of an inch, a number of feet, hundreds of feet, or even several 
miles. In the Plateau region of Arizona and Utah, several faults of 

Fig. 233, A pivotal or Huge fault. The fault runs from left to right, between tilted 
rocks and steep cliff on nearly horizontal strata. Displacement decreases toward the 
right, increases toward the left. Maximum throw about 3000 feet. Spotted Range, 
Nevada. (Longwell.) 

great magnitude extend in a north-south direction, some of them cross- 
ing the Grand Canyon. Each of the largest fractures in this group cai> 
be followed 100 miles or more, and has a throw measured in thousands 
of feet. The Great Basin region presents the phenomenon of faulting 

on a colossal scale. In the 
area between the Sierra Ne- 
vada on the west and the 
Wasatch on the east, the 
crust is divided into huge 
blocks by gigantic fractures; 
and differential displacement 
of these blocks, together with 
erosion, has resulted in 
mountainous topography. A 

sunken tract of country due to downfaulting, or to uplift of adjacent 
areas, is called a graben (German for trough or ditch). Illustrations are 
the Jordan Valley and the Dead Sea, the great Rift Valley of Africa, 
and Death Valley in eastern California. An upstanding mass between 
two fault troughs or grabens is a horst. 

Thrust Faults. Reverse faults are most common in those regions 
where crushing and folding of the Earth's crust have taken place; and 
the stronger the folding or crushing has been, the greater and more 

Fig. 234. A normal fault passing into a mono- 
clinal fold. 


evident the reverse faults are. They are especially evident in the strati- 
fied rocks of mountain regions, as in the southern Appalachians and in 
the Alps. Careful and detailed study of old eroded mountain areas has 
disclosed reverse faults of tremendous displacement, some of them with 
a comparatively low angle of dip, or even quite horizontal. A reverse 
fault that has a gently inclined fault surface is known as a thrust fault, 
or simply a thrust. Many of these are of such magnitude and impor- 
tance that they are commonly considered by themselves as a special 
class of faults. The surface on which movement occurs is spoken of as 
the thrust surface, or less accurately as the thrust plane. 

Such thrusts have been discovered and studied especially in the Alps, 
in northwestern Scotland, in the Scandinavian peninsula, in the southern 
Appalachians, in the Rocky Mountains from British Columbia to Utah, 
in southern Nevada, California, and in many other regions. The hori- 
zontal displacement of lower, older f ormations over younger rocks ranges 
from several miles up to 25, 30, or even 40 miles for individual thrusts. 
Figure 235 represents a portion of the great thrust along the front ranges 

Lewis Range 

Fig. 235: Section, showing the thrust in northern Montana, whereby very old geo- 
logic formations of the pre-Cambrian are made to override the much younger beds of the 
Cretaceous. BB is the surface of thrusting; D, J>, and Chief Mountain are erosional rem- 
nants of the pre-Cambrian resting on, and surrounded by, the younger Cretaceous. Dis- 
placement by thrusting observed, 7 miles; total amount unknown. (Generalized after 

of the Rocky Mountains in northern Montana. The deciphering of 
these great displacements is one of the triumphs of modern geological 

Topographic Results of Faulting. If a fault of considerable mag- 
nitude were to be formed suddenly, it would naturally be marked by 
displacement of the Earth's surface, giving rise to a cliff, or scarp. 
Numerous fault scarps are recognized, and some of them have been 
formed within historic time. 

Such scarps in their original form may be called initial fault scarps. 
As the process of weathering and erosion works more actively in general 
on the uplifted side, the scarps become dissected and lowered, and slowly 
retreat from the fault line. Thus they pass through youthful, mature, 
and old stages. Finally, the difference in elevation on opposite sides of 
the fault line may disappear completely, and thus all topographic ex- 


pression, initially due to faulting, may be obliterated. This would 
finish one cycle of erosion on a faulted surface (Fig. 236). 

If now the whole region should be uplifted, without further displace- 
ment on the fault, and thus a new cycle of erosion initiated, then the 


Fig. 236. Shows the origin, development, and possible history of an initial fault scarp. 
A, block of strata containing two harder, more resistant intruded sheets of trap, before 
displacement. B, after faulting and some erosion; the fault scarp has become mature, 
and has retreated from the fault line. C, approaching the end of the first cycle of erosion; 
the fault scarp has been obliterated. 

agents of erosion might find on opposite sides of the fault line rocks 
of quite different hardness and ability to withstand their attack. There- 
fore one side might be lowered so much more rapidly than the other as 
to leave the latter standing in a cliff or escarpment (Fig. 237). As such a 
cliff would result, not directly from the initial faulting movement, but 
from subsequent differential erosion, it deserves a distinguishing name, 
and has been termed by W. M. Davis a fault-line scarp. (See Fig. 237, 
which continues the history of the fault shown in Fig. 236.) The varying 


Fig. 237. Possible development of fault-line, scarps. A, the faulted block of the 
preceding figure commencing a second cycle of erosion; intruded trap sheets more resistant 
than the enclosing beds; uplifted block to the right. B, after erosion; a fault-line scarp 
has formed which faces toward the uplifted block. (7, continued erosion has carried away 
the top trap of B and a new cliff has formed facing the other way, toward the sunken 
block; this fault-line scarp resembles the initial fault scarp (compare B, Fig. 236). 

resistance to erosion on the opposite sides of the fault line determines 
naturally on which side the scarp will form, and a very long and old 
fault line might be marked in different parts of its course by cliffs 
facing in opposite directions. Finally, through the completion of 
another cycle of erosion these cliffs in turn might be worn away. The 


influence of an old fault that reaches deep into the crust 'is evident in the 
topography so long as erosion of the area is active. Determination of 
the exact history of any fault is possible only after very careful geological 

The east slope of the Sierra Nevada and the west slope of the Wasatch 
Mountains are fault scarps that have undergone a large amount of ero- 
sion. Nevertheless the faulting is recent in a geological sense. At 
the base of each of these ranges some of the movement has occurred so 
recently that initial scarps may be seen almost uneroded in the soft 
fans of alluvial material brought down by the streams. The last re- 
corded movement near the Sierra Nevada took place in 1872. These 
faults have grown by successive movements, and the upper part of each 
scarp has suffered most from erosion. 

The Plateau region, through which Colorado River cuts its way, is 
crossed by a series of great faults which are marked by prominent cliffs. 
These have been described as actual fault scarps; but most of them are 
fault-line scarps developed during a second cycle of erosion. Examples 
on a smaller scale are very common. Thus the sunken tract of sand- 
stone and intercalated trap sheets between New Haven, Connecticut, 
and Springfield, Massachusetts, is divided into a series of tilted blocks 
by faulting. It has passed through at least one cycle of erosion, in 
which the initial fault scarps were eroded away; it is now in another 
cycle, initiated by broad uplift of the region without further movement 
on the faults. 

Many old faults with vertical displacements amounting to thousands 
of feet are now practically unrecognizable in the surface forms. Ob- 
viously this relation in each case indicates erosion of great magnitude. 
Either there was a high fault scarp which slowly wasted away or the 
growth of the displacement has been so slow that erosion has kept up 
with it. This last suggestion is not altogether unreasonable, for we can 
scarcely im'agine that the formation of great faults, with miles of dis- 
placement, has been a sudden process. Rather it results from gradual 
yielding of the shell of the Earth to forces brought to bear upon it during 
long periods of time. 

The detection of faults that do laot show any distinct topographic 
relief is possible through several kinds of evidence. The most common 
and obvious is the disturbance, or discontinuity, produced in the struc- 
ture of the rocks, especially the stratified formations. In homogeneous 
masses of igneous rocks the recognition of faults is more difficult; yet 
even here discontinuity in certain features, such as dikes and veins, may 
lead to the discovery of faults and furnish a basis for measuring the dis- 


Origin of Faults. -AThe immediate cause of faults is comparatively 
simple and generally agreed upon; they result from strains set up in the 
outer shell of the Earth.) Relief occurs by movement of rock masses, 
either along the surfaces of some previous fracture, or by the formation 
of a new one. Compressive strains give rise to reverse faults and thrusts; 
and it is natural that features of this kind occur commonly in areas of 
folded rocks, as folds also represent failure under compression. In 
regions of broad warping the rocks may be broken by torsion or twisting, 
which sets up tensional strains. After fracturing, displacement occurs 
by gravitative settling and readjustment of the fault blocks. Thus over 
wide regions where the strata are not otherwise disturbed, as in the 
Colorado Plateau, they are penetrated by fractures on which there have 
been great displacements. Also in the upper portions of uparching folds 
there may be tension and cracking, with subsequent gravitative settle- 
ment and faulting. 

The ultimate cause of faulting evidently depends on those processes 
within the Earth which give rise to compressional or tensional forces 
and so set up strains in the lithosphere. These forces are most strikingly 
displayed in the formation of its chief features of relief, such as moun- 
tains and plateaus; and faulting may be considered only an attendant 
result of their operations. The forces themselves are hidden and can 
be inferred only from their effects. As the subject is obscure at best, 
and speculation must be guided by consideration of all available facts, 
it is best to postpone inquiry into the ultimate cause of crustal deforma- 
tion until the structure and history of mountains have been discussed. 


Definition. It is not uncommon to find, on examining the stratified 
rocks exposed in cliffs, valleys, and mountain sides, that one set of beds, 
whose composition, parallel position, and contained fossils prove them to 
be a continuously deposited series, rest upon another set of rocks, whose 
position and characters show equally well that they were formed at an 
earlier period and under other conditions. Thus in the diagram, Fig. 238, 
the layers of strata d have been deposited at one period and under one 
set of conditions; they are, therefore, spoken of as a conformable series of 
beds. Also beds of the series c are conformable among themselves; but 
it is quite evident that they are not conformable with d. Their attitude, 
and the abrupt termination of each bed upward, indicate that this series 
was tilted strongly and subjected to erosion before deposition of the 
overlying strata. The two series are unconformable with respect to one 
another, and the surface db separating them is called an unconformity. 



It should be understood clearly that the unconformable contact is a 
widespread surface and not merely the line exposed in a vertical section. 
Such a contact represents approximately an 
old land surface, or an old wave-scoured sea 
floor, and therefore is a cIearjOf^T)f erosion. 
It is not essential thaFthe rocks beneath an 
unconformity be stratified. The lower forma- 
tion might be composed of igneous rocks, such 
as granite; or of metamorphic rocks, such as 
schist or gneiss. If the surface ab can be Fi - 23S - ~ Section to show 

. -, ,.~ i ... , , r . unconformity. The figure rep- 

identifaed positively as a record of erosion pre- res ents a vertical outcrop such 
ceding the deposition of overlying strata, it as might be seen in the wall of a 

r j. T . _ , canyon. A conformable set of 

IS a SUrlace Ot Unconformity. If the Contact strata d rest uneonformably 

between rock masses of unlike character is due u P n another conformable set 

ri . . . . c; the line a6 represents the un- 

to lauiting or to igneous intrusion, the contact conformity. / 

surface is not an unconformity. 

Geologic History Revealed. Suppose that c and d, Fig. 239, are 
two unconformable series of marine strata. Then the relationships 
represented in the diagram indicate a definite succession of geological 
events. First there was a long period of quiet deposition in which the 
beds of set c were laid down in a horizontal position, or nearly so, on a sea 
floor. The thickness of the beds, the kinds of rocks (limestones and 

shales), and the contained 
fossils constitute the record 
for this period. At some 
time later than the depo- 
sition of the youngest beds 
in this series, there was 
strong folding or faulting 
by which the strata were 

tilted steeply. Possibly the deformed mass reached to mountain heights. 
At any rate the deformation was succeeded by erosion, which planed the 
upended strata to a nearly even surface. From the section shown in the 
diagram we have no means of estimating the duration of uplift and ero- 
sion. Not only were there no records of the time formed in sediments 
in this area, but those of the previous period were wasted and obscured. 
Therefore there is a gap or "lost interval" in the geological record at this 
locality. Next in the geological history there followed a period of sub- 
sidence, when the eroded surface became sea bottom again, and received 
a new deposit of sediments, forming the conformable series of strata d. 
The events of this time are recorded continuously in the strata as before. 
Finally, after a second period of uplift, tilting, and erosion, the whole 

239. An angular unconformity with both 
series of beds tilted. 


record is presented to us to be read so far as the evidence permits. The 
history here given may then be summarized as follows: first, deposition 
of strata; second, tilting, elevation, and erosion; third, subsidence and 
fresh deposition; fourth, final elevation, with tilting and erosion. If 
another subsidence should occur in the near future, the present land sur- 
face would mark a second abrupt break, above which a new set of hori- 
zontal strata would be deposited. 

Relation to Bonds of Rocks. Since an unconformity at the base of 
marine strata represents a submerged land surface, certain kinds of rocks 
are naturally associated with it. The sea advances inland, as a result of 
land submergence and of its own ceaseless gnawing at the shoreline. 
Where the land and sea meet there is generally a beach of the ordinary 
type, and- as the land subsides this beach marches inland at the edge of 
the encroaching s6a. Every part of the newly made sea bottom will 
have been passed over by this advancing beach; and all the superficial 
cover of the land soil, pebbles, and rock decayed by weathering is 
worked over by the advancing sea, and converted into beach material. 
The finer particles of the ground-up detritus are swept out to sea, and 
only the gravel and sand remain in the agitated waters near the shore. 
As an end result of this process, a continuous layer of conglomerate or 
coarse sandstone the old beach material commonly lies directly 
above the unconformity. The coarseness and thickness of this deposit 
varies according to the kind of rock composing the old land, the rate at 
which the sea has advanced, and other factors. A basal conglomerate or 
sandstone does not invariably accompany unconformities. More rarely 
the lowest deposit of the new marine series is shale or even limestone. 

Classification of Unconformities. Unconformities may be divided 
into two main groups. In the first, the lower formation, either by the 
tilting of the beds or by its composition of non-stratified rocks, shows at 
once its nonconformity with the series of beds above it. It is called an 
angular unconformity (Figs. 240, 241, 242, C and D) if the bedding planes 
of two stratified series meet at an angle. This term is good so far as it 
goes, but it does not cover the whole case, since the lower formation is 
not always composed of stratified rocks but may be of massive igneous 
or metam orphic rocks (Fig. 242, A and B). A more general term is 
needed, and an unconformity of this class is here termed a noncon- 

On the other hand, the lower formation may be elevated, eroded, and 
submerged without material disturbance of the position of the beds. The 
old and the new formations will then have their stratification planes 
actually, or practically, parallel. This constitutes an unconformity of 
the second class, and as it is desirable that it should be distinguished from 



Fig. 240. Angular unconformity between tilted and eroded Paleozoic limestones and 
horizontal Tertiary conglomerate. Meadow Valley, Nevada, along the Union Pacific 
R. R. (Longwell.) 


Fig. 241. A closer view of the unconformity pictured above, in another part of 
Meadow Valley. (Longwell.) 




E F 

Fig. 242. Diagrams to illustrate various kinds of unconformity. A, sedimentary 
strata deposited on metamorphic and igneous rocks; B, nonconformity on massive igneous 
rocks; C, angular unconformity between two sedimentary series, the later series undis- 
turbed; D, angular unconformity, both series tilted; E, disconformity, both series of 
strata horizontal; F, disconformity, both series tilted together. 

the other, it has been termed a disconformity (Fig. 242, E and F). We 
have then the following cases of unconformity: 


1. Nonconformity, two formations with visibly different structure, 
a. Lower formation of rocks nonstratified, or apparently so. 

6. Lower formation of stratified rocks, tilted with relation to 
overlying strata (angular unconformity). 

2. Disconformity, two formations in parallel position separated by 

erosion surface. 


Obviously the subject of unconformity is closely connected with the 
study of sedimentary rocks; but a full appreciation of its meaning re- 
quires some understanding of crustal disturbance. Strong angular un- 
conformities record severe deformation, either by folding or by faulting; 
and widespread disconformities indicate warping movements that have 
involved large areas. A study of unconformities emphasizes the close 
relationship between crustal movements, erosion, and sedimentation. 
These ancient records, like the present wasting surface of the land, rep- 
resent the continual struggle between deep-seated forces that produce 
irregularity, and surface processes that strive to keep the lands feature- 
less and low. 


1. Structural and Field Geology; by James Geikie. 426 pages. D. Van Nostrand 
Co., New York, 1905. 

Describes the common types of structural features, with examples drawn largely 
from Europe. 

2. Field Geology; by Frederic H. Lahee. 607 pages. McGraw-Hill Book Co., 
New York, 2nd edition, 1923. 

Gives excellent descriptions of structural features in general, with particular 
emphasis on types of structure that affect economic geological work. 


It is difficult to think of earthquakes apart from their relation to human 
affairs. From the earliest recorded times the recurrent shaking of the 
solid ground, with consequent destruction on the surface, has been a 
cause of terror to man. Repeatedly, and in widely separated localities, 
populous communities have suffered great loss of life and property. De- 
structive earthquakes recorded during the brief span of human history 
are numbered by thousands. Geologic evidence indicates that violent 
shocks have been recurrent throughout the history of the Earth; and 
there is every reason to expect their frequent occurrence in the future. 

The serious aspect of earthquakes from the human viewpoint is realized 
on review of some major catastrophes. September 1, 1923, approxi- 
mately 100,000 lives were lost as a result of the Tokyo earthquake, and 
the estimated property loss exceeded $4,000,000,000. The shocks at 
Messina in 1908 and at Kansu, China, in 1920 were equally disastrous to 
life. According to report, more than a million and a half persons were 
killed by ten Chinese shocks between the eleventh and twentieth cen- 
turies; and Mallet, a profound student of seismology (from seismos, an 
earthquake) estimated that for the whole Earth at least 13,000,000 lives 
were lost through earthquakes in the course of 4000 years. Some 
activities of .man himself, such as wars, or the operation of automobiles, 
result in a much higher death rate. Yet earthquakes are especially 
productive of fear, probably in part because they come without warning, 
and in part because their cause is more or less mysterious. Study of 
earth shocks from a geologic standpoint has dispelled a part of this 

Cause of Earthquakes. An earthquake is a trembling, or undulatory 
motion, in the more or less elastic rock shell of the Earth, communicated 
to it by an impulse or shock of some kind just as a bell is set in vibration 
by a smart tap on its side. The shock or impulse is evidently the im- 
mediate cause of the earthquake; but what is the origin of such shocks? 
Ancient philosophers who sought to explain natural* phenomena by 
natural causes generally connected earthquakes in some way with the 
weather. Aristotle and Lucretius thought they were produced by winds 
rushing out of the Earth and leaving voids, with consequent collapse. 
Modern scientific evidence shows that shocks may arise from several 




causes, most of which must be considered of minor importance compared 
with one major source, which appears to give rise to all great earthquakes. 

One minor cause is in violent volcanic outbursts, like that of Krakatoa 
in 1883 and of Bandaisan in Japan in 1888; but earthquakes produced in 
this way are light in intensity and quite limited in extent. Moreover, 
many outbursts are not attended by any shocks, or at best by only feeble 
tremblings, such as occurred during the eruption of Mont Pelee in 1902. 
For a long time it was thought that volcanic action was an important 
source of earthquakes, and this idea is frequently revived; but the careful 
comparison of the two phenomena, especially in Japan, has shown that 
there is no necessary connection in occurrence between heavy earthquakes 
and volcanic eruptions. Instruments near Kilauea, in Hawaii, record 
numerous minor tremors sometimes hundreds of them in a single 
month. Very few of these are accompanied by visible volcanic activity, 
although it is probable that shifting of magma at some depth is the prin- 
cipal cause of the local shocks. 

Another minor cause of earthquakes may be the sudden caving in of 
subterranean cavities, or collapse of their roofs under the weight of 
superincumbent rock masses. This is most likely to happen in regions 
underlain by limestone, 
since large quantities of this 
rock are removed in solution 
by underground waters. It 
is possible, as has been sug- 
gested, that the earthquakes 
which in 1811 devastated 
the lower Mississippi valley, 
especially about New Ma- 
drid in southern Missouri, 
were partly due to this 
cause; though the area 
affected is so extensive and 
the effects of the earthquake 
shocks were felt to such 
great distances that caving 
probably was not the prin- 

cipal Cause. Fig. 243. Map of a part of California, showing 

T+ "hocj -nnw hppn rathpr the position and extent of the fault line, A~A, move- 

_LL Hcto ULUW kJCCMJ. -Lcno-LCA 

ment along which produced the earthquake of April 

definitely settled that most is, 1906. 

of the major earthquakes 

result from "the jar given by sudden yielding to strain in the Earth's 

crust. Such yielding may be by formation of a new fracture, or by 


abrupt displacement along the walls of an already existent fault. In 
many areas visited by disastrous shocks the surface of the ground has 
been broken along fault lines and the amount of displacement is clearly 
indicated. Commonly these movements take place along old fault 
zones which bear the marks of repeated displacement. In California a 
great fracture zone can be followed almost continuously, by means of its 

Fig. 244. Trace of the fault concerned in the California earthquake of 1906. The 
deep soil above the bedrock broke irregularly. As movement was horizontal, no scarp 
was formed at this locality. (U. S. Geol. Surv.) 

peculiar surface expression, from the southern part of the state north- 
westward for 600 miles. This feature, known as the San Andreas Rift, 
passes near the city of San Francisco (Fig. 243). On April 18, 1906, 
abrupt movement along at least 270 miles of this fracture caused a de- 
structive earthquake. The length of this break is somewhat exceptional 
among historical earth movements; but similar breaks 25 to 50 miles long 
are not uncommon. 
A careful study along the San Andreas Rift, after the rupture in 1906, 


yielded valuable information on the nature and amount of the dispkce- 
oaent. In general no scarp was made by the faulting, because the motion 
was almost entirely horizontal, parallel to the fault (Fig. 244). This fact 
was established beyond question by the offsetting of roads, fences, and 
other features that extended across the break. Some roads were cut 
sleanly across, and offset considerably more than their width (Fig. 245). 
The largest measured displacement was 21 feet. More commonly a part 

Fig. 245. Horizontal displacement of a road by movement on the San Andreas Rift 
in 1906. Before the earthquake the two offset portions of the road were in a straight line. 
The fault extends from left to right, directly across the road. (U. S. Geol. Surv.) 

of the motion on a break of this kind is vertical and results in a steep 
scarp (Fig. 246). 

If it is considered that the walls of a fault are pressed closely together, 
that movement is possible only by overcoming great f rictional resistance, 
and that the displacement, once it occurs, takes place almost instantane- 
ously, it is not surprising that powerful vibration is set up in the vicinity 
of the fault line. The exact nature of movement along the San Andreas 
Rift was made the subject of special study, and it is concluded that mass 
movement in the crust on opposite sides of the fault was in slow progress 
for years before 1906. Deformation in the rock was by bending, until 
the strain could be borne no longer and relief occurred by abrupt slipping 
along the old fracture. According to this idea, the sharp movement of 



1906 was confined to a comparatively narrow belt closely adjacent to the 
fault, and did not involve the immediate shifting of great segments in the 
crust. The jar, resulting from the release of energy that had been ac- 
cumulating for decades, was in effect a heavy blow which made the 
Earth tremble. 

It is not to be supposed that a visible fault appears in every area visited 
by an earthquake. Commonly the direct evidence of crustal movement 
is wanting, especially in connection with mild or moderate shocks. It is 

Fig. 246. Displacement on a fault at Midori in the Neo Valley, Japan, in 1891. The 
former plain was broken, and one part dropped with relation to the other. Note vertical 
as well as horizontal offsetting of the road. (K. Ogawa.) 

believed that actual displacement occurs very frequently at considerable 
depth and does not reach to the surface. This is a logical inference, as 
every break must be limited in extent, vertically as well as horizontally. 
Earthquakes of the first rank, however, are in a general way restricted to 
regions of active faulting, for which there is evidence at the surface. 
Some destructive shocks originate under the sea; but even in this event 
it may be evident from soundings that faulting has displaced the sea 
floor. For example after the Tokyo earthquake of 1923 it was found 
that an area in Sagami Bay had dropped more than 1000 feet below its 
former depth. 

It was thought at one time that earthquakes were generated from a 
point at some depth below the surface, and this was called the focal 


. 341 

point, or centrum. The point immediately over this on the surface was 
called the epicenter. This latter point was determined by drawing con- 
centric closed curves, called coseis?nal lines, on a map of the region 
through points of simultaneous arrival of the waves, 
as indicated by clocks (Fig. 247). By other mathe- 
matical methods the distance below the epicenter of 
the focal point was calculated. These methods led 
to discordant results for some earthquakes, and 
eventually to the discovery that for any one earth- 
quake there might be several epicenters situated in a 
line, or that where earthquakes habitually occurred in 
a given region the different epicenters were situated 
along a line. Such a line probably represents a fault, 
even if there is no surface evidence of its existence. Fi 947 _ M of 
The terms centrum and epicenter still have value, coseismal lines. Black 
although it should be understood that they are not Angles represent 

J points at which the 

points. time of the earthquake 

Effect of Shock. It is important that we distin- 7 a a t * ly recorded accu " 
guish clearly between cause and effect in earthquake 
phenomena. The displacements shown in Figs. 245 and 246 are not the 
results of earthquakes as is commonly supposed; they represent the 
causes. The effect of the sudden movement along a fault is to set up 
vibrations that move outward from that place, and these constitute the 

earthquake, as it is perceived at a 
distance. Thus the earthquake is 
propagated as a series of waves in 
the highly elastic body of the Earth. 
When these elastic waves emerge at 
^ the surface the loose ground is 
thrown into rapid vibration, ordi- 

Fig. 248. Wire model showing path nar ily with an amplitude not ex- 
traveled by a particle of matter during an . , T , , , 

earthquake; after Sekiya. ceeding a few inches. In bedrock 

the amplitude is only a fraction 

of an inch. The actual amount of movement reaches a maximum 
in deep alluvium that is saturated with water. On terrane of this kind 
destructive effects are greatest. For example at San Francisco the 
devastation was most acute on the low alluvial fiat near the bay. 
Buildings on solid rock, even much nearer the fault, suffered less damage. 
The contrast in behavior of the deep alluvium and of bedrock may be 
illustrated by striking, sharply a bowl in which there is jelly. The bowl 
is set vibrating with resulting sound, but actual motion in the walls of 
the vessel is imperceptible. However the same impulse transmitted to 


the jelly sets up longer waves that are visible. Excessive destruction on 
wet alluvium is caused by this larger wave motion as compared with the 
behavior of solid rock. In any kind of material the motion is not simple 
and rhythmic, but very complex (Fig. 248). 

Recent Examples. On August 31, 1886, the city of Charleston, 
South Carolina, was visited by a severe earthquake which did great 
damage. The shock was distinctly felt as far away as Chicago, a dis- 
tance of 800 miles. This shock is of special interest, because from general 
considerations it does not appear that severe crustal disturbance should 
be expected at Charleston. 

In 1899 a great earthquake took place in southern Alaska. As the 
region is mostly uninhabited the shock passed almost without notice at 
the time. Studies which have since been made show that considerable 
alterations in topography took place at the time of its occurrence, espe- 
cially about Yakutat Bay. Marked changes were also induced in the 
great glaciers of this region by the shattering of the ice and by snow- 
slides from the mountains. 

In August, 1906, the coast of Chile was visited by a severe earthquake, 
which did great damage in Valparaiso and other places. After-shocks 
continued for a long time while readjustment along the fault was going 
on. The west coast of South America is noted for its earthquakes, in 
connection with which notable elevation of the coast line has occurred. 

One of the greatest disasters in modern times occurred on Dec. 28, 
1908, when Messina and Eeggio, cities on the narrow strait which sepa- 
rates Sicily from the mainland of Italy, were completely destroyed by a 
terrific shock. Evidently the area is in a zone of crustal weakness and 
readjustment, as severe earthquakes have occurred repeatedly. 

The great Tokyo earthquake of 1923 is remarkable for the very large 
changes in topography that were produced by the crustal disturbance. 
Parts of the shore around Sagami Bay were lifted up as much as 6 feet. 
In addition to the large depressions on the floor of this bay, other parts 
of the floor were lifted several hundred feet, making shoals where there 
had been deep water. 

These are only a few examples out of many that might be selected. 
Scarcely a day passes that shocks are not recorded from some part of the 
world by earthquake observatories. 

Seismic Belts. Although earthquakes occur in all parts of the world, 
they are most likely to happen in certain well-defined tracts, which lie in 
the two great seismic belts. One of these follows the western coast of 
North and South America, the Aleutian Islands, and the island groups 
along the eastern coast of Asia, and thus borders the Pacific Ocean on the 
east, north, and west. The other includes the Mediterranean, the Alps, 



the Caucasus, the Himalayas, and continues into the East Indies, where 
it intersects the first belt at a large angle. (Figs. 249 and 250.) In a 
general way these zones coincide with the great volcanic belts (page 259) ; 
and this fact might appear to support the idea that volcanoes are an 
important cause of earthquakes. However, since the belts correspond 

Fig. 249. Map of seismic belts in the Eastern Hemisphere. On S. L. Penfield's 
stereographic projection. (Compare map of volcanic belts, Fig. 181.) 

closely to young mountain systems and other marks of recent crustal 
movement, it is probable that both earthquakes and volcanoes have a 
common cause in this disturbance of the lithosphere. It is a notable 
fact that where the seismic belts lie directly along the continental bor- 
ders, as on the coast of Chile and the eastern coast of Japan, the land 
descends sharply, without any broad intervening shelf, to great depths 
of the ocean. Some of these steep slopes descend into foredeeps, which 
are great troughs that appear to be sinking, while the bordering lands are 
rising. We conclude that these are zones of weakness in the Earth's 


crust where strains are being constantly relieved by movements, and in 
which, therefore, earthquakes are continually recurring. 

It is commonly thought that certain regions are practically exempt 
from danger of earthquakes because no real disaster has happened in 
them since they have been settled and cities have sprung up. It is true 

Fig. 250. Map of seismic belts in the Western Hemisphere. On S. L. Penfield's 
stereographic projection, (Compare Fig. 182.) 

that most of the Atlantic coasts, and large areas in continental interiors, 
are relatively free from earthquakes. The comparative stability around 
the Atlantic as compared with the Pacific is emphasized not only by the 
historical record, but by the existence of a wide continental shelf, which 
is in strong contrast with the Pacific foredeeps. However, the experi- 
ences of New Madrid in 1811 and of Charleston in 1886 are a warning 
that no locality may be entirely exempt. Even in New England, which 
is not recognized as a seismic tract, there has been an average of one 
perceptible tremor a year since the settlement of the country. Probably 



none of the shocks has been of maximum intensity, although several have 
caused some destruction. 

Submarine Earthquakes ; Tsunamis. The location of seismic belts 
suggests that many earthquakes originate under the ocean. Their oc- 
currence beneath the sea is shown by shocks communicated to vessels on 
the surface above, and by rupturing of submarine cables. Since the 
invention of sensitive instruments by which it is now possible to record 
distant earthquakes and determine their location, it has been learned 
that a large proportion of all earthquakes occur on the floor of the 
Pacific. The most conspicuous mark of a submarine earthquake is the 
huge wave that commonly is generated in the ocean by disturbance of the 
floor. Such waves have long been known as tidal waves, a misleading 
name since they have no connection with the tide. They are now gener- 
ally known to seismologists by their Japanese name tsunamis; or they 




? _ _ < ... 











.SDI^flQ 3q 






x: +'*- Ki5ir : 

jSSd J 








^iv. _ * jgj *C 





.. . . . ^-.^ _j- 



12A.M. 3 AJt. 4A.M. 6A.M. 8AJU. 10A.M. 12 Noon 

2P.M. 4PJC ttfcjt. SPJL 10P.H.12AJL 

Fig. 251. Record of tsunami by tidal gauge. Vertical lines represent time spacing 
on the paper, driven horizontally by clockwork. Horizontal lines show height in feet as 
recorded by the rising and falling pencil of the gauge. 

may be called seismic sea waves. Some are of immense size, measuring 
100 or even 200 miles from crest to crest, and as much as 40 feet high. 
They are so broad that in the open sea they are not ordinarily perceived; 
but on approaching the coast they may pile up in huge breakers and, 
sweeping far inland, cause enormous damage and loss of life. 

Lisbon in 1877, Japan in 1854 and in 1896, Peru in 1868, suffered from 
great and disastrous tsunamis. The number of victims of a single 
inundation of this kind has been as great as 20,000. These vast waves 
are felt over whole oceans and move with tremendous speed, from 300 
to 500 miles per hour. Those from Japan have crossed the Pacific in 
about 12 hours. At such distances their height may be only a few inches; 
but the ebb and flow of from 15 to 30 minutes, like small subordinate 
tides, are registered as wavy lines on the record of a tidal gauge (Fig. 251). 
These records make it possible to determine the size of the wave, since 
they give the period of oscillation, and since the velocity can be cal- 
culated from the time and location of the shock that caused the tsunami. 



Recording Earthquakes. Very delicate instruments have been 
invented, called seismographs (Fig. 252), which record the tremors due to 

distant earthquakes; and the study 
of these records has led to impor- 
tant geological conclusions. The 
principle upon which the common- 
est instruments are constructed is 
simple. If a heavy mass of metal 
be suspended like a pendulum, ow- 
ing to its inertia it will remain for a 
time at rest when the shock arrives, 
while the bedrock vibrates beneath 
it. A pencil of some kind is se- 
cured to the suspended weight, 
and rests lightly on a paper or other 
medium suitably prepared to record 
the motions of the pencil. When 
the bedrock oscillates the heavy 
weight acts essentially as a steady 
" -- - - - - point; but the vibration is trans- 

mitted to the pencil, and may be 

Fjg. 252. Diagrammatic representation ^o^fi^ ^ rnanv fimpq fl<* rlp^ir^rl 

of a seismograph. The upright post (G) is niagmned as many times as aesirea 

attached firmly to bedrock. The heavy by a Simple mechanical device. If 

weight (above M) is connected with the post ,, rian^r I'nqfpflr? of heiTio- made 

only by a freely moving joint (L) and a flexi- tne P a P 6r > l&Steac^ OI Dem maae 

ble wire (D) . Records are made by the stylus fast, be a Strip continuously Carried 

1 Whi h " alongly clockwork, thepencil when 

at rest will draw a straight line 
upon it; when vibrations of the Earth occur the line will bend sinuously 
from one side to the other (Fig. 253). Such a record is known as a 
seismogram. Some instruments have, instead of a point or pencil, a 
small mirror that throws a beam of light upon a sheet of photographic 
paper. The seismogram is revealed only after the sensitized paper is 

Although the principle of a modern seismograph is simple, in con- 
struction some of the instruments are rather complicated since they are 
arranged to record not only horizontal motion in two components, but 
also the vertical motion as well. It is from such records in three di- 
rections that the wire models like that shown in Fig. 248 are constructed. 
Since the intervals of time are marked on the moving paper, the instru- 
ment records the time of arrival of the shock and also the duration. 
The directions of diversion of the markers from their regular paths show 
also the direction from which the shock has come. 



Seismograms. The study of seismograms of distant earthquakes has 
led to the discovery that the main shock is preceded by smaller rapid - 
vibrations which are recorded when the seat of disturbance is several 
hundred miles or more from the recording station. These are known as 


Fig. 253. Record of the earthquake in Messina on Dec. 28, 1908, as shown by a seis- 
mograph in Gottingen, Germany, over 1000 miles distant. The actual vibration ex- 
perienced by the instrument is greatly exaggerated in the seismogram. 

the preliminary tremors. Thus a normal seismogram has the characters 
seen in Fig. 254. It has been determined that these preliminary tremors 
represent elastic impulses that come by the shortest path through the 
Earth; that is, in the general direction of a chord from the seat of dis- 
turbance to the recording station; whereas the later large vibrations 
represent those elastic waves that have traveled by a longer route over 

i i 

Fig. 254. Seismogram of distant earthquake; ab, first preliminary tremors; bc t 
second preliminary tremors; ce, main shock; fh, later phases; hi, tail. (After Omori.) 

the surface circumference. The first preliminary tremor (a, Fig. 254) is 
caused by a compressional or longitudinal wave (commonly known as the 
primary), which travels several miles per second; the other preliminary 
(6, Fig. 254) represents a transverse wave motion (the secondary wave), 
which travels at a distinctly slower rate. Therefore the time interval 


between the two preliminary tremors is proportional to the distance 
traversed, and from this information the distance between the seat of the 
shock and the seismograph can be calculated accurately. By computing 
the distances from at least three separate stations, and drawing circles on 
a map with these distances as radii, the circles will intersect in a common: " 
point, which is the locus of the earthquake. 

It is obvious that the circumferential or long waves will move out from 
the locus in opposite directions on any great circle of the Earth. - If the 
locus of the shock should be exactly on the opposite side of the Earth 
from the seismographic station, these two sets of long waves would 
reach the instrument at the same time. Ordinarily one of the arcs is 
much longer than the other, and the second set of long waves arrives 
much later than the first; or if the distance is very great and the shock is 
slight, the second set may die out before reaching the instrument. Most 
seismograms record a succession of vibrations following the first long 
waves. Some of these later phases represent recurrent after-shocks, and 
others are due to complex reflected wave motions in the Earth. 

Geological Deductions from Seismograms. The fact that the pre- 
liminary tremors, which are supposed to travel through the Earth, arrive 
at distant points so long a time ahead of the main shock, cannot be ex- 

plained alone by the shorter path traveled. 
The time interval shows that they are also 
propagated at a much greater rate of speed 
than the vibrations traveling in the outer shell 
of the Earth. The deduction from this is 
that they move in a more elastic medium than 
the superficial part of the crust. Moreqver, 
the concordant results in different directions 
show that inside of the outermost layer, 
which we know is heterogeneous in composi- 
tn- orr -n ^ rx tion, the Earth is homogeneous, or regularly . 

Fig. 255. Paths of transmis- ; & } & J 

f earthquake shock through arranged around its center in structure; or, 

*' if ^homogeneous, the heterogeneous parts 
are relatively so small and numerous that 
different paths of considerable length through them give the effect of 
uniformity. Furthermore the average velocity increases with the dis- 
tance of the recording station; thus the average rate of transmission 
along sz, Fig. 255, is greater than along sy, which in turn is greater than 
along sx. Velocities of the primary and secondary waves, as calculated 
from available earthquake data, vary with depth as shown in the fol- 
lowing table: 



Depth Below Surface, 

Velocity, in Miles per Second 

in Miles 

Primary Wave 

Secondary Wave 















These results show not only that velocity increases down to a certain 
depth, but more and more slowly as the depth increases, and this would 
seem to indicate that the density and elasticity of the Earth increase 
with depth down to a certain region. It is a matter of considerable 
importance that the rate of propagation actually diminishes below a 
depth of nearly 2000 miles; and at still greater depth the secondary wave 
is not transmitted. Seismographs whose distance from the locus of an 
earthquake exceeds one-third the Earth's circumference (120) receive 
no record of the transverse wave. The chord connecting the ends of an 
arc of one-third the circumference cuts the Earth's radius at its middle 
point. There is strong indication, therefore, that the Earth has an inner 
core, about 4000 miles in diameter, whose composition or state is different 
from that in the shallower zones. As will be explained in a later chapter, 
there is good reason for believing that the core consists chiefly of metal 
instead of rock. Some scientists have suggested also that this metallic 
core may be in the fluid state. This would explain the failure of the 
transverse waves to penetrate the core, as such waves are transmitted 
only through elastic solids. 

From the fact that the rate of speed increases with the depth in the 
outer 2000 miles, it follows that the quickest path of wave transmission 
from the seat of shock to a distant station in this portion of the globe will 
not be a straight line, as from s to y in Fig. 255, along the chord of the arc, 
but will be a curved line slightly concave upward, somewhat like the line 
scy. In other words, by following this line the waves gain more in time 
in entering more elastic layers than they lose in distance, and hence 
seismologists generally assume that the path followed by the waves 
making the pre1i.miTifl.ry tremors at a distant recording station is curved. 
This is of some importance because, assuming the path to be straight 
and noting the fact that the preliminary tremors do not generally show 
in seismograms unless the distance is greater than 600 miles, the 
deduction has been drawn that there must be a rather sharp boundary 
between an outer rocky heterogeneous shell of the Earth and an inner 


homogeneous core, and that, since the chord of an arc of 600 miles at 
its middle point is 12 J miles below the surface, this must be the thick- 
ness of the outer layer. If a curved path is assumed, the thickness must 
be considerably greater. But Reid has suggested that the probable 
reason the preliminary" tremors do not show in the records of "near" 
earthquakes is that instruments are not generally delicate enough to 
record and distinguish them from the principal shock, until distance 
produces appreciable time intervals. This view, expressed several years 
ago, is strengthened by recent investigations. Sensitive modern instru- 
ments situated less than 100 miles from epicenters have differentiated 
the preliminary tremors. 

Geological Effects of Earthquakes. There are several geological 
effects from earthquakes, but they are, comparatively speaking, of minor 
importance. The loose mantle of soil and other debris is often ruptured 
by the passage of the wave with the formation of fissures, which may be 
of some depth. A more important effect is the starting of landslides and 
avalanches in mountainous regions, through the jarring of the Earth. 
A variation in the flow of water from s'prings, or even the forming of new 
springs, has also been observed. 

Much more important are the movements of the crustal blocks at the 
time of earthquakes; but as previously emphasized, these are the cause 
and not the effect of the shocks. 


1. Our Mobile Earth; by R. A. Daly. 342 pages. Charles Scribner's Sons, New 
York, 1926. 

Chapters I and II give an excellent discussion of earthquakes and their cause. 
The style is vigorous and stimulating. Numerous excellent illustrations. 

2. A Manual of Seismology; by Charles Davison. 249 pages. Cambridge 
University Press, 1921. 

A clear presentation of principles and methods in earthquake study. 

3. Report on the California Earthquake of April 18, 1906; by A. C. Lawson and 
others. 451 pages. Publication No. 87, Carnegie Institution of Washington, Vol. 1 
and Atlas, 1908. 

A full account of the San Francisco earthquake, with a description of the great 
San Andreas fault. Numerous excellent photographs and maps. 


Definition of Metamorphism. In addition to the igneous and sedi- 
mentary rocks previously described there is a third class termed the 
metamorphic, whose distinctive characters are due to metamorphism. 
These rocks are the records of the remarkable transforming power of 
certain geologic forces that are at work within the Earth's crust. 

Metamorphic means changed in form, and metamorphism is a general 
term for all those changes by which the original characters of rocks are 
more or less thoroughly altered, so that the component minerals or tex- 
tures of the rocks are transformed into new minerals or textures, or both. 
Some metamorphic rocks pass gradually into those whose fossils and 
stratification prove them to be undeniably of sedimentary origin, whereas 
other metamorphic rocks grade into rocks whose characters show con- 
clusively that they are of igneous origin. Prom these transitions we learn 
that some metamorphic rocks have been formed from sedimentary and 
some from igneous rocks. As we shall see later, some metamorphic rocks 
have also been derived from other metamorphic rocks. 

The metamorphic changes may be so profound that the resultant prod- 
uct no longer resembles the rock from which it was derived but has 
become a new rock. Sedimentary rocks thus thoroughly metamorphosed 
are more coarsely crystalline, and the fossils that they may have contained 
and even the marks of stratification have been completely obliterated. 
Igneous rocks also, if severely metamorphosed, have lost their original 
distinctive features. 

Limestone, for example, may be metamorphosed into coarse-grained 
marble with consequent loss of color and obliteration of the fossils that 
were in it; basalt may be converted into a green, slaty rock that gives no 
hint of its original igneous nature. Rocks that represent the stages of 
transition between the limestone and the marble or between the basalt 
and the green slate can be found; but under metamorphic rocks we in- 
clude only those which have been so profoundly changed that their 
original outward characters have been either entirely obliterated, or 
nearly so, and distinctly new rocks have been formed. We say that it is 
the "outward" characters which are obliterated, for the bulk chemical 
composition of a rock as a rule remains unchanged during metamorphism. 



The various changes that rocks undergo from the effects of weathering 
might in the strict etymologic sense of the term be classed as metamor- 
phic. But they have been already discussed under the work of the 
atmosphere and the production of soils; therefore these agencies and the 
weathered rocks and soils that result from their action are not considered 
in this place. 


Three kinds of metamorphism are recognized: 1, contact metamor- 
phism; 2, dynamic metamorphism; and 3, load metamorphism. Contact 
metamorphism is produced by the action of intrusive igneous masses on 
the rocks into which they were intruded. The effects produced are due 
chiefly to the heat supplied by the magma and the hot gases that issued 
from it when it consolidated. These effects are of course limited to the 
vicinity of the contact with the igneous mass, and hence the term contact 
metamorphism; but the term igneous metamorphism, by emphasizing 
the agency that produces the changes, is more appropriate. The most 
obvious form of contact metamorphism, probably the most impressive 
to the layman because coming more nearly within the ken of ordinary 
experience, is the conversion of a coal bed into a layer of coke by an 
injected sill. 

Dynamic metamorphism results from the action of tangential pressure 
in the Earth's crust, as displayed in folding of the strata and in great 
overthrust faults. As metamorphism of this kind often accompanies 
profound dislocations of the crust, it is called by Heim, the great master 
of metamorphic geology, dislocation-metamorphism. 

In places rocks appear to have been metamorphosed without either 
the intervention of igneous masses or of dynamic metamorphism. Here 
the cause of the metamorphism was apparently that the rocks were for- 
merly deeply buried, having become depressed in the crust under a heavy 
load of overlying strata. The increase in temperature brought about by 
the deep subsidence has caused the development of new minerals and 
textures, and the heavy pressure due to the load favored the production 
of heavy minerals, such as garnet, which require the minimum volumes. 
As the Earth's own heat is the main factor here in bringing about the 
transformations, the process is sometimes called geothermal meta- 

The best established example of metamorphism of this kind is fur- 
nished by the German potassium-salt deposits. These salts were laid 
down in an evaporating arm of the Permian sea; they were many and of 
complex compositions; they were stable, however, under the conditions 
of moderate temperature then prevailing. Subsequently the basin in 


which they were deposited slowly subsided and they became covered by 
20,000 feet of sedimentary beds. The temperature of the deeply buried 
salt beds rose to that determined by their depth in the crust, and drastic 
rearrangements in the composition of the salts took place and many new 
minerals were produced. 

Silicate rocks are far less sensitive than the potassium-salt minerals. 
Nevertheless, certain formations of pre-Cambrian age are believed to 
have acquired their metamorphic condition by load metamorphism. In 
places also the bottoms of synclines may have become so deeply down- 
folded that they have been subjected to load metamorphism. 

Just as elsewhere in Geology, hard and fast lines do not separate the 
various kinds of metamorphism. Contact metamorphism may go on 
concomitantly with crustal folding; but we shall omit the description of 
these complex phenomena and devote our attention chiefly to the simpler 
forms of contact and dynamic metamorphism*. 

We shall begin with the contact-metamorphic rocks, for their origin is 
well established. As to the other metamorphic rocks, sometimes known 
as the crystalline schists in lieu of a better name, it is not always clear 
whether they are of dynamo-metamorphic origin or of load-metamorphic 
origin, or whether they were formed by the cooperation of contact 
metamorphism with dynamic or load metamorphism. 


As previously explained, the term contact metamorphism is used to 
denote the changes that are induced in rocks by the intrusion into them 
of a mass of magma. 

The most noticeable effect of contact metamorphism is a baking, har- 
dening, or toughening of the intruded rocks in a zone that surrounds the 
intrusive igneous mass. Since the intensity of these changes diminishes 
with distance from the intrusive mass, the contact-metamorphic zone is 
sometimes called somewhat figuratively a contact aureole. 

Width of Contact Zone. The width of the contact-metamorphic 
zone depends chiefly on the size of the igneous mass. The widest zones 
occur around stocks and batholiths. Around them the contact zone may 
be a mile wide, or even more; usually it is some hundreds of yards wide, 
but adjacent to a small intrusion such as a dike, it may be only a few feet. 
Lava flows produce at most a slight baking of the soils or rocks on which 
they rest. 

The width of the contact zone around one and the same igneous mass 
may vary, for it is controlled by the configuration of the igneous body and 
by the attitude of the surrounding rocks. Thus in Fig. 256, section 1, as 



a result of the lesser slope of the contact a wide zone is produced at CD, 
much wider than that adjacent to the vertical contact at AB. And in 
section 2 the beds at F } which slope toward the igneous rock, tend to have 
their bedding planes opened, and to furnish easy passageways to the 
emanations from the cooling magma. Since these emanations are the 
chief agents in carrying the heat and producing the metamorphism, it is 
clear that a broad zone F will be made on this side, compared with E, 
where conditions are reversed and a narrower zone must be formed. 

Fig. 256. Sections of intruded stocks and their contact zones. In 1, the breadth on 
the surface CD is greater than AB, depending on shape of igneous mass. In 2, the width 
P is greater than in E, depending on inclination of beds. 

Contact Metamorphism of Rocks of Different Kinds. The extent 
and the intensity of contact metamorphism depend very much, in ad- 
dition to the factors already discussed, on the kinds of rocks surrounding 
the igneous mass. For our purpose here the sedimentary rocks may be 
divided into the three groups: sandstones, shales (and clays), and lime- 
stones. On pure sandstone the effect is rather small, though near the con- 
tact the sandstone may be changed into quartzite a compact rock so 
firmly cemented that it fractures across the grains, instead of around them. 
Limestone is changed into marble, the masses of which may extend for 
considerable distances from their contact with the igneous rocks. Shales 
show the most notable and, generally, far-reaching results. The soft 
shales are greatly hardened, and near the contact they are converted 
into a rock known as hornfels, which to the unaided eye strongly resem- 
bles a black fine-grained igneous rock, such as basalt. 

Coarse-grained igneous rocks, being the products of magmatic con- 
solidation and therefore having already been at high temperatures, are 
generally but little affected by later intrusions. However, where vol- 
canic rocks, such as basalts, have been invaded by granite batholiths, 
extensive and drastic metamorphism is produced. 

The most interesting results are produced in limestones, especially 
impure, cherty varieties. Not only are the limestones turned into mar- 
ble, but a great variety of minerals are newly formed in them, depending 
on the reactions that take place between the bases 'and acidic oxides 
present, especially lime and silica. Thus when the limestone is heated 


above 500 C. the silica tends to drive out carbon dioxide, 
CaC0 3 + Si0 2 = CaSi0 3 + C0 2 

and calcite is changed into calcium silicate (wollastonite). If the limp. 
stone contains dolomite, then the following reaction may occur: 

CaMg(CO 3 ) 2 + 2 Si0 2 = CaMg(SiO 3 ) 2 + 2 C0 2 

and a pyroxene is formed, and carbon dioxide liberated. It is found thai 
pyroxene is thus formed in the inner, hotter portion of the aureole 3 
whereas an analogous compound, a white amphibole (tremolite), is 
formed in the outer, cooler portion of the aureole; thus these two minerals 
can be made to serve in a rough way as geologic thermometers. 

If other impurities occur in the limestone, such as clay furnishing 
alumina, and iron oxides, many other new minerals will be formed, 
These more complex reactions may be illustrated by the following 

Calcite -f Clay -f Quartz = Garnet + Carbon dioxide + Water. 

3 CaC0 3 -f H4Al 2 Si 2 9 + SiOa = CasAlaSisOi* + 3 COa +2 H 2 O. 

Thus a limestone that contained clay and sand as impurities may be 
changed into garnet with evolution of carbon dioxide and water. 

Normal and Pneumatolytic Contact Metamorphism. In the normal 
contact-metamorphic zone the effects produced are entirely owing to the 
heat given off by the igneous mass: the surrounding rocks become 
heated to a high temperature and new minerals are produced by the 
recombination of the elements already present in them, and consequently 
the chemical composition of the rocks remains unchanged. A marble; 
for example, has the same chemical composition as the limestone from 
which it was formed. If, however, the igneous mass during its consoli: 
dation gives off gases carrying iron, silicon, boron, etc., vastly different 
results are produced in the surrounding rocks. The most striking results 
are produced in limestones, because of the readiness with which lime- 
stones react with the magmatic gases. Many new minerals are formed, 
often beautifully crystallized. The resultant contact-metamorphic 
rock differs greatly in composition from the original rock: substances 
have been added in large amount by the gases that streamed from the 
magma. As gases are the means by which this transfer is effected and 
the reactions produced, contact metamorphism of this kind is termed 
pneumatolytic. If the gases contained iron, copper, or tungsten in 
notable quantity, valuable ore deposits may be formed in the limestones 
as incidental by-products of metamorphism of this kind. 



Most metamorphic rocks have not been formed by contact metamor- 
phism, at least under the static conditions that prevailed while the rocks 
of the hornfels type were produced. So well known is this fact that the 
term metamorphic rocks when used without qualification refers to the 
great group that is not of contact-metamorphic origin. We lack a good 
distinctive designation for this group, and for want of a better name they 
are called the crystalline schists, in virtue of two characteristic features 
that most of them have : their obviously crystalline appearance and their 
foliated or schistose texture. 

Many of these crystalline schists are the products of dynamic meta- 
morphism, during which the rocks from which they were derived were 
forced by heavy differential pressure to flow in the solid state. As a 
result of this flowage there were developed either new textures, or new 
minerals, or both. Examples of such rock flowage are well shown in the 
Alps, perhaps the most marvelously folded tract in the world, where the 
limbs of the folds have become attenuated to one-tenth, even to one- 
hundredth of their original thickness and the arches of the folds have 
become enormously thickened: manifestly solid material has flowed 
from the flanks of the folds to the points of flexure. 

Minerals of Metamorphic Rocks. The minerals in rocks differ 
greatly in their ability to withstand the changes of temperature and 
pressure to which different metamorphic processes subject them. When 
new chemical and physical factors operate on them, they tend to change 
into new minerals, which are stable under the new conditions. The result 
of this adjustment is a metamorphic rock. The igneous rocks are char- 
acterized by a distinctive set of minerals, chiefly silicates, whereas car r 
bonates and hydrated oxides as well as silicates are abundant in sedi- 
mentary rocks. Some minerals like quartz have a wide range of stability 
and occur in all three classes of rocks, but many minerals when subjected 
to metamorphic processes are converted into other minerals; thus car- 
bonates are likely to be changed into silicates. Quartz, feldspars, mica, 
and hornblende occur in both igneous and metamorphic rocks, whereas 
garnet, staurolite, kyanite, talc, chlorite, and serpentine occur chiefly 
in metamorphic rocks. 

Foliation. Most metamorphic rocks are visibly crystalline to the 
unaided eye, and they have a parallel arrangement of their constituent 
minerals. This parallel arrangement gives the rock a foliated texture 
(from foliumj a leaf), and a rock having it is known as a foliate. By 
reason of it a rock tends to split, or cleaye, more or less perfectly into 
flakes or slabs parallel to the foliation, A coarsely foliated rock is called 


a gneiss, and one in which the foliation is well developed and closely 
spaced is termed a schist. Although foliation is the characteristic tex- 
ture of metamorphic rocks, there are a few, such as serpentine, marble 
and quartzite, ^hich commonly show no trace of this texture. In some 
rocks the parallel texture is straight or nearly so for considerable 
distances, as seen in Fig. 257; commonly, however, the banding is very 
much contorted, bent, or curled, showing the amount of deformation to 
which the original rocks were subjected (Fig. 260). 

The foliated texture is due to the occurrence of the component min- 
erals in separate layers or flat lenses, or to the parallel arrangement of 
prismatic or tabular minerals, such as hornblende or mica, or to a com- 
bination of both modes of arrangement. It is a result of the granulation 
and recrystallization to which the original rocks were subjected, and has 
been imposed on igneous and sedimentary rocks alike. The resemblance 
of the banding or lamination of schistose rocks to stratification led in the 
past to the erroneous view that all of them were derived from stratified 
rocks; that some have also been made from igneous rocks was learned 
much later. 

Slaty Cleavage. The cleavage exhibited by metamorphic rocks is 
most remarkably developed in slates^ so characteristically that this 
variety of it is called slaty cleavage. Slates used for roofing, blackboards, 
and other purposes show this cleavage. The question of its origin has 
stimulated much investigation, along both experimental and mathe- 
matical as well as geological lines. From these studies it has become 
clear that the cleavage is the result of great pressure on the material, and 
that the plane of cleavage is at right angles to the direction of pressure that 
produced it. When fine-grained sediments, such as muds and clays, are 
subjected to intense pressure, the oblong particles in them tend to rotate 
so that their lengths are perpendicular to the direction of pressure; they 
also tend to become flattened perpendicularly to it/ A ;Important also is 
the fact that many of the platy or elongated minerals, such as the micas 
and the chlorites, have an excellent 'cleavage parallel to the flat direc- 
tions. A considerable part of the minerals in the slate, such as the 
micas, were not originally present in the sediments, but were formed 
during metamorphism accompanying the pressure, and as- they grew 
they set themselves in parallel orientation. All these features tend to 
give the rock a capacity to cleave readily in one direction. 

The cleavage planes do not necessarily bear any fixed relation to the 
bedding planes. The beds were laid down in horizontal position and the 
direction of pressure is also horizontal or nearly so. Therefore the 
cleavage planes, being developed at right angles to this pressure, may cut 
the bedding at highly inclined or right angles. For the beds may become 



Kg. 257^ Banded gneiss, Portland Township, Ottawa Co., Quebec. 
(Geol. Surv. of Canada.) 



folded before the pressure becomes intense, and the cleavage planes, being 
developed after the folding, will consequently intersect the bedding at 
various angles, though these 
cleavage planes themselves are 
all strictly parallel to one an- 
other (Figs. 258, 259, and 261). 
Although most slates have 
been made from fine sediments, 
such as muds and clays, slaty 
rocks have also been produced 
by the shearing of fine-grained 
igneous rocks, such as felsites _. co 01 , __ _ 

, , , . , 1 - , Fig. 258. Slaty cleavage (represented by the 

and basalts, and beds OI VOl- nearly vertical lining) in folded beds. 

canic ash. In the making of 

slates the original characteristic features of the rocks may become 
greatly distorted and even obliterated; thus fossils and pebbles in 
the stratified rocks and embedded crystals and other structures of 
the igneous rocks may be flattened into lenses or squeezed out into 

Kg. 259. Slaty cleavage cutting at a high angle beds folded in a syncline. 
Slatington, Pa. (U. S. GeoL Surv.) 

cylinders. Cleavage may in places be mistaken for original bedding, 
unless care is taken, and consequently the geological structure may be 
wrongly interpreted. It is sometimes important to indicate the cleav- 
age on geologic maps, and this may be done by plotting its dip and 
strike, like that of a bedding plane. The important relation that cleav- 


age bears to mountain ranges and their origin will be discussed under 
that subject. 

Places of Occurrence. Crystalline schists are widely distributed 
over the Earth's surface, and in some regions they are the only rocks 
exposed over extensive areas. Such extended tracts of metamorphic 
rocks were formerly said to be due to regional metamorphism 3 but inas- 
much as that term explains nothing it is becoming obsolete. Its chief 
merit is its noncommittal character: it leaves open the question whether 
the metamorphism is due to igneous, dynamic, or geothermal activity, 
or to combinations of these activities. 

Metamorphic rocks are particularly abundant in the most ancient 
(pre-Cambrian) formations the world over. In the United States they 
occur quite generally in New England, in the Adirondacks, and in a 
strip of country running from New York to Georgia. New York, Phil- 
adelphia, Baltimore, and Washington are built on metamorphic rocks. 
Metamorphic rocks are well shown in the inner gorge of the Grand 
Canyon of the Colorado, and in many other places. 

The metamorphic rocks form also the cores of many mountain ranges, 
in which they have become exposed by denudation. The structure of 
these mountain ranges, as will be discussed later, includes folded strata, 
and the degree of metamorphism of the rocks is proportional to the close- 
ness and intricacy of the folding. In other ranges, however, strata just 
as closely and intricately folded have not been metamorphosed. Hence 
mere intricacy of folding does not determine metamorphism. Probably 
the depth at which the folding takes place within the crust and the speed 
with which it occurs are the controlling factors. A relatively rapid rate 
of folding would generate the heat necessary to effect metamorphism. 
- The relation between folding, metamorphism, and mountain formation 
is so common that, where we find rocks intricately folded and very meta- 
morphic, we assume that highlands once existed but have been eroded 
away, or, in general, that metamorphic rocks are exposed to our view 
only as a result of deep erosion. In conformity with this idea, metamor- 
phic rocks are regarded as of continental origin, because they imply 
(when of sedimentary origin) the following sequence of events: deep 
erosion of a land mass to supply sediment; the deposition of this sedi- 
ment in beds; the folding of the beds to effect metamorphism, perhaps 
with incidental production of a mountain range; and lastly renewed 
erosion to expose the metamorphic rocks. Such an array of processes 
can occur only on a great scale and therefore on and about continents; 
consequently when metamorphic rocks are found in place on Fiji, New 
Caledonia, South Georgia, and other islands, it is held that this proves 
that these islands are really remnants of continental masses. 


Age of Gneisses and Schists. Some of the facts previously men- 
tioned led to the view that gneisses and schists must be geologically very 
ancient. This conclusion is now known to be but partly true. For 
although nearly unmodified beds of early geologic age that have been 
changed but little from their original horizontal position occur in Russia 
and in the upper Mississippi valley, on the other hand comparatively 
recent strata that have been greatly folded, such as those in the Alps and 
in some other mountains, have been strongly metamorphosed. The 
metamorphic condition of rocks depends wholly on whether or not they 
have been subjected to metamorphic processes; consequently the older 
the rocks are, the more likely it is that in the vast span of geologic time 
they have been affected by them: that is the large kernel of truth in the 
older view, 


Introductory. Because the metamorphic rocks are derived from 
igneous, sedimentary, and also from metamorphic rocks, it follows that 
there must be an extraordinary diversity of metamorphic products. 
Yet in the same way that we were able for ordinary purposes to gather 
the igneous and stratified rocks into a few groups, so we can consider the 
metamorphic under the few most important types. 

Many sedimentary rocks are largely made of the detritus derived from 
disintegrated igneous rocks. In the process of disintegration it may hap- 
pen that there is not much weathering and chemical change. Conse- 
quently the resultant sedimentary deposit will not differ much from the 
original igneous rock in chemical and mineral composition. Thus the 
red-brown sandstone (arkose) of the Connecticut valley, which contains 
much feldspar, has practically the same composition as the granite of the 
adjacent region. If such arkoses should become so thoroughly metamor- 
phosed as to lose their original characters, they could not be distinguished 
from metamorphosed granites, nor could their former status be deter- 
mined. From this example it will be clear that, while we can tell the 
origin of some metamorphic rocks at once, as is true of marble and 
quartzite, and can ascertain the origin of others after careful study in the 
field and laboratory, we are unable to ascertain the origin of many 
metamorphic rocks. 

Ckssification. It is possible to show in a general way the relation 
between the most common sedimentary rocks and their metamorphic 
derivatives in the following table: 




Sedimentary Rocks 

Metamorphic Rocks 



Gneiss, and various schists 



Quartzite and various schists 

Silt and clay 


Slate, phyllite, and various schists 

Calcareous deposits 


Marble, and various schists 

The igneous rocks, it will be recalled, are roughly divided into two main 
groups, the one chiefly composed of light-colored feldspathic minerals, 
and the other mostly of dark ferromagnesian minerals. We can illus- 
trate in a rough way the relation between them and their metamorphic 
derivatives in the following table: 

Igneous Rocks 

Coarse-grained feldspathic types, such as 
granite, etc 

Fine-grained feldspathic types, such as 
felsite, tuff^ etc 

Ferromagnesian rocks, such as gabbro and 

Metamorphic Derivatives 

Slate and schists 

Hornblende schistSj various schists, 
and serpentine 

Comparison of the tables will show that gneisses and schists may have 
diverse origins, as previously pointed out. Combining the results of 
these tables, we obtain the following main groups of metamorphic rocks, 
distinguished according to their mineral composition or by their texture, 
or by a combination of both. 


-1. Gneiss, coarsely foliated rock. 

,2. Quartette. 

3. Mica schist. 

4. Slate and phyllite. 

5. Hornblende schist; talc and chlorite schists. 
15. Marble; mixed carbonate-silicate rocks. 

, 7. Serpentine. 

Gneiss. The most common variety of gneiss (pronounced nice) 
consists, like granite, of quartz, feldspar, and mica, but as the mica is 
arranged in more or less parallel planes, the gneiss has a rude cleavage. 
Some hornblende may accompany or replace the mica, and other min- 
erals, such as garnet, may also occur, giving different varieties. Gneisses 
range in color from light to dark, and from fine to coarse in grain. Va- 
rieties transitional between granite and gneiss are very common. In 



those gneisses made from conglomerates the original pebbles may still 
show as lenticular masses. 

Gneiss is one of the most common of metamorphic rocks, and its 
many varieties have been formed under very diverse conditions: some 
under conditions of mild dynamic metamorphism in which the minerals 
have been merely mechanically deformed, and others under the condi- 
tions of most intense metamorphism, in which the mineral composition 
has been wholly reconstituted. 

Some granite gneisses are called primary gneisses, because the foliated 
or gneissic texture has been assumed during magmatic fiowage or by the 

Fig. 260. Contorted gneiss; Fullerton, Hudson Bay, Canada. 
(Geol. Surv. of Canada.) 

shearing of magmas still in a partly consolidated or viscous condition. 
Batholiths that have been intruded during orogenic activity generally 
have foliated borders of primary gneiss. A view of beds of gneiss is 
seen in Fig. 260. 

Quartzite. Quartzite is a rock composed of quartz grains which are so 
firmly cemented that fracture takes place through the grains, instead of 
around them. Originally the rock was a sandstone and has been changed 
to a quartzite either (1) by filling the pore space of the original sandstone 
through the deposition of quartz from circulating ground water, or (2) 
by contact metamorphism, as already explained, or (3) by dynamic 
metamorphism. Quartzites formed as the result of the deposition of a 
quartz cement by ground water are not regarded as metamorphic rocks. 


Quartzites of dynamo-metamorphic origin are commonly interbedded 
with gneisses and mica schists. The original pore space in the sandstone 
has been eliminated mainly by compaction and rearrangement of the 
quartz already present in the rock. Although quartzites of this origin 
are intimately associated with foliated rocks, they themselves rarely show 

Quartzites are generally compact hard rocks of light colors white, 
gray, reddish, or buff and are likely to be of vitreous appearance. 

Mica Schist Mica schist is the most widely distributed and im- 
portant member of the great class of crystalline schists. The essential 
minerals are quartz and mica, and it is especially the mica that gives the 
rock its distinctive character. Different varieties of mica occur; the 
most common is a silvery white muscovite, which gives the rock a bril- 
liantly spangled appearance, and the black mica biotite also is 
common. The micas are in irregular leaves or tablets with their cleavage 
planes oriented in the direction of the schistosity, and it is in fact this 
parallel arrangement of the micas that produces the extraordinary fissile 
character of mica schist. Some mica schists are dotted with dark-red 
garnets in well-formed crystals. 

Mica schists are rocks that have attained a high-grade metamorphic 
condition. The initial material from which they were derived was an 
argillaceous sediment a shale of some kind and the transformation 
of the dull, amorphous substance of the shale to a brilliantly spangled 
mica schist is in a way as remarkable as the metamorphosis of a chrysalis 
into a beautiful butterfly. 

Slates. The origin of slates from fine-grained sediments, such as 
muds, clays, and ash deposits, by the action of compressive forces has 
already been discussed. While they may have various colors red, 
green, gray, etc., the most common one is dark gray to black, due to 
carbonaceous material in the original muds. As is well known, they are 
quarried for roofing slates, blackboards, and other purposes (Fig. 261). 
They are closely related to shales, but the distinction between them is 
that the cleavage or fracture of a shale is conchoidal, or "shelly" (whence 
-its name, shale), and is parallel to the bedding planes; whereas in slates 
it is a secondary induced structure, which, as previously stated, is not 
necessarily parallel to the bedding. 

Slates are metamorphic rocks of very low grade; some indeed the 
clay slates are very feebly metamorphic and are argillaceous sediments 
that have merely assumed a metamorphic texture, i.e., the slaty cleavage; 
others the mica slates have acquired new minerals as well as the 
slaty cleavage. 

Phyllites. Phyllites resemble slates, but their constituent mica is 



coarser, which gives them a silky, glimmering luster. Most of them are 
transitional between slate and mica schist and are the results of an inter- 
mediate grade of metamorphism : more intense than necessary to produce 
a slate and not sufficiently intense to produce a schist. Some, like the 
ordinary slates, have been formed from sediments, but are more highly 
metamorphic. Others have been made from igneous material felsite 
lavas, tuffs, etc., by shearing and accompanying agencies of meta- 

Hornblende schist. Hornblende schist is a common and typical 
variety of schist. It is generally dark green to black, and the parallel 

Fig. 261. Illustrates the occurrence of slates and cleavage. Slate quarries, Browns- 
ville, Me. (U. S. Geol. Surv.) 

prisms of hornblende, if not too large, usually give it a silky luster. 
Talc schist and chlorite schist are other common schists, in which talc 
and chlorite are respectively the predominant minerals. 

Marble. Marble is the metamorphic equivalent of the sedimentary 
carbonate rocks, such as limestone and chalk. Generally the marks of 
bedding and the fossils are effaced during metamorphism and the mate- 
rial is converted into grains of calcite that are visibly crystalline to the 
unaided eye. It is, therefore, harder, more compact, with purer colors, 
and takes a good polish. Just as there are ordinary limestones consisting 
only of calcium carbonate, and dolomitic limestones containing calcium- 
magnesium carbonate, CaMg(CO 3 ) 2 , in variable quantity in addition to 


the CaC0 3 , so we have caltite marbles and dolomite marbles. Com- 
mercially this chemical difference is not important, but geologically it is 
of interest because the kinds of minerals that are likely to occur scattered 
more or less thickly through them are quite different in the two. 

Marble is generally massive and shows no cleavage, even when it has 
been subjected to great stresses. As rocks go it is relatively plastic and 
flows under moderate pressures. If for example a dike or bed of schist 
is inclosed in marble that is forced to flow under differential pressure, the 
dike or bed, being brittle, will be ruptured and torn apart, and the marble 
will flow in between the dissevered fragments. 

Pure marble is white, and the mottling, banding, and colors shown 
by ornamental varieties is due to impurities; the red and yellow tones to 
oxides of iron, the grays and blacks to varying proportions of organic 
matter. Besides being produced by dynamic metamorphism, marble is 
also formed by contact metamorphism. 

Serpentine. This name is given to a mineral, a hydrous silicate of 
magnesium, H 4 Mg 3 Si 2 09, and also to a rock largely or entirely composed 
of it. The rock is usually greenish to black, soft, of a greasy feel, and 
massive, or without cleavage. Some of the blotched, lighter-green 
varieties are used as building and ornamental stones. Most serpentines 
appear to have been made by hydrothermal metamorphism (action of 
hot waters) on deeply buried masses of igneous rock rich in magnesia, 
such as peridotite, whereby the magnesium silicates change to this 
hydrated variety. Some impure dolomite marbles contain magnesium 
silicates (olivine, pyroxene, etc.) which may alter to serpentine. Some 
verde antique appears to be a serpentine-bearing marble of this nature. 


Metamorphic rocks, such as the series beginning with slate and com- 
prising phyllite, mica schist, and garnet gneiss, represent stages in 
progressive or advancing metamorphism. Each successive rock in the 
series is the product of higher-grade metamorphism than the one that 
precedes it, chiefly as the result of adjustment to conditions of pro- 
gressively higher temperature. A rock that has become adapted to the 
condition of highest metamorphic intensity may, however, be subse- 
quently shifted into a new geologic environment, in which the conditions 
of stability are those of lower-grade metamorphic intensity. A garnet 
gneiss, for example, may be reduced to a phyllite along the base of a 
great overthrust block of the Earth's crust. Outwardly the resultant 
phyllite resembles a phyllite produced by progressive metamorphism, 
but internally, as shown by the microscope, it gives evidence of its former 


high-rank metamorphic condition. By such retrogressive metamor- 
phism many varieties of metamorphic rocks have been produced from 
other metamorphic rocks. By this process the already astonishing 
diversity of metamorphic rocks is greatly increased. 


1. The Principles of Petrology; by G. W. Tyrrell. 349 pages. E. P. Dutton and 
Company, New York, 1927. 

Part III of this volume gives the only modern presentation in English of the 
difficult subject of metamorphism. 

2. Metamorphic Geology; by C. K. Leith and W. J. Mead. 337 pages. Henry 
Holt and Company, New York, 1915. 


Knowledge gained by study of the rocks at the Earth's surface impels 
man to speculate about the hidden interior. The materials and condi- 
tions that exist at great depth can never be known by direct observation. 
Openings in the solid Earth, such as mines and deep wells, are very super- 
ficial in comparison with the long radius of the globe. We infer that 
some of the rocks now exposed in the cores of old mountains were at a 
depth of several miles before erosion laid them bare; but even so they 
were always a part of the "outer shell/ 7 and we cannot be sure of their 
exact nature while they were under the relatively light load that has 
been removed in the course of long ages. If it were possible to make and 
maintain an opening to the center of the Earth, what would be revealed? 
Would the composition of material be found to change radically with 
depth, so that a large part of the interior would bear little or no re- 
semblance to the superficial rocks? What would be the temperature at 
various levels? Would any large part of the interior be in a liquid 
condition? How would the materials at great depth behave under the 
enormous overburden? These are profound questions. They are not 
prompted by idle curiosity merely; they represent the goal of scientific 
investigations whose attainment would make clear many of the phe- 
nomena seen at the Earth's surface. 

Whenever direct evidence is not available, science turns to the indirect 
and circumstantial. It is possible, with aid of this sort, to draw certain 
inferences that are sound; but if we seek to advance farther into the 
unknown, we must be content with hypothesis and speculation. Some 
suggestions can be accepted as probabilities, with the understanding that 
they may be found wanting after further investigation. It is well, at the 
outset, to state clearly the actual basis of fact, and to outline the methods 
of attacking the problem. In this way we shall avoid misconceptions, 
and be able to evaluate each suggestion on its merits. 


Size and Shape of the Earth. The science of geodesy, which is con< 
cerned with exact measurement and mapping of the Earth's surface, has 
determined with precision the dimensions and the form of the globe. 



This information, obtained by great labor through cooperation of scien- 
tific men in many countries, is of fundamental importance in problems 
relating to the Earth as a whole. It is known that the equatorial 
diameter exceeds the length of the polar axis by nearly 27 miles. As this 
figure is about ^ of the diameter, it is stated that the ellipticity of the 
Earth is yj^. This departure of the globe from a true sphere is largely a 
response to rotation. Important deductions may be drawn from this 
known distortion of the Earth, considered with relation to other facts. 

Gravity and Density. By precise physical experiments we determine 
the "constant of gravitation." The method involves essentially the 
measuring of the force with which the Earth attracts a body of known 
mass. The result gives the total weight of the Earth; and since its size 
also is known, it is a simple matter to compute the average density. This 
value, arrived at by many experimenters, is 5.52; that is, an average 
sample of the Earth weighs five and a half times as much as an equal 
volume of water. 

Direct determinations of density, using all kinds of rocks known at 
the surface, give an average value of 2.7. As this is less than half the 
density of the whole Earth, it appears that the interior must consist of 
much heavier material than the outer part. Other deductions will be 
discussed in a later paragraph. 

Relation Between Density and Form. In detail, the surface of the 
solid Earth is highly irregular. Continents stand well above ocean 
floors; plateaus, mountains, and deep troughs make irregularities of 
smaller order. On a small globe made to true scale these surface fea- 
tures appear insignificant; but in actual proportions some of them are 
large, and from a human viewpoint the irregularity is of the utmost 
importance, as without it the oceans would be world-wide. 

Theoretically the surface of the Earth would be quite smooth if the 
material below the surface were uniform in its character. In reality, we 
know that the rocks 'exposed to observation differ considerably in com- 
position and in density. Basalt and other dark-colored igneous rocks 
are appreciably heavier than granite. By geological investigation and 
by highly technical instrumental determinations, a strong probability 
has been established that the rocks underlying the oceans are in general 
denser than those composing the continents. Moreover it is concluded 
from careful geodetic study that the great mountain ranges of the Earth 
are composed of or underlain by material that is slightly deficient in 
density compared with the crust as a whole. It would appear from these 
facts that the larger surface irregularities are not haphazard, but have a 
fundamental cause; that differences in surface elevation are compen- 
sated by differences in rock density. Many important inferences and 


some theories are based on this relationship. These matters will be 
discussed in a later part of the chapter. 

Behavior Toward the Moon and Sun. The moon and sun exert a 
constant pull on the Earth, and the yielding of the oceanic waters to this 
pull gives rise to the tides. By careful and ingenious experiments it has 
been determined that the solid body of the Earth also responds to the 
tidal force. However, the yielding is very minute : about what would be 
expected if the Earth were composed of the strongest steel. 

Another effect is produced by lunar and solar attraction on the equa- 
torial bulge of the Earth, causing the globe to wabble slowly as it spins. 
Consequently the north pole of the heavens shifts slowly from year to 
year. This effect, called precession, is known precisely, and the forces 
involved can be calculated closely. The use of this information will be 
mentioned later. 

Response of the Earth to Seismic Waves. The transmission of 
earthquake vibrations through and around the Earth has been discussed. 
By study of the through-waves, the elastic properties at various depths 
can be deduced. In general, the Earth reacts to these impulses as a 
highly rigid, elastic body. The transverse or distortional wave in par- 
ticular is of great significance, as this type of elastic wave is transmitted 
only through solid material. It is also significant that the velocity of 
both the transverse and the longitudinal waves decreases below a depth 
of about 2000 miles. This change appears to indicate either a different 
kind or a different state of material in the central core. 

Other Facts About the Earth. Volcanoes and hot springs indicate 
high temperatures at depth locally, and direct measurements in deep 
mines and wells suggest a universal increase in temperature downward. 
The rate of this increase varies between wide limits. In some places it is 
as high as 1 F. in 30 feet; in other places, as in the deep gold mines of the 
Transvaal, it is only 1 F. in 250 feet. The average for all observations is 
about 1 F. for 60 feet. Thus in wells that go to a depth of a mile and a 
half, the difference in temperature between top and bottom may be 125 
F. or more. Some mines that reach down several thousand feet are 
uncomfortably warm. 

The determination in recent years that all known rocks contain ap- 
preciable quantities of radioactive material is of great importance. It 
appears certain that the distribution of these materials must be confined 
to a comparatively shallow outer zone. Radioactive elements such as 
uranium and thorium break down at a constant rate, producing heat; and 
as rocks conduct heat away at an extremely slow rate, it can be calcu- 
lated that the existence of these elements aven at moderate depths would 
in permanent fusion. Granites contain higher percentages of 



Fig. 262. Rock fiowage in laboratory test by Adams and Bancroft, (a) longitudinal 
section through steel cylinder (CC) with rock column (R) in place. Part of the steel wall 
was reduced to small thickness, as shown, (b) same after slow application of great pres- 
sure (up to more than 100,000 pounds per square inch) on the two pistons, (c) a rock 
column after and before the test. 


radioactive substance than other known rocks, and hence it is argued 
that granites, which are the predominant rocks in the continents, are 
limited to shallow depth. 

Behavior of Rocks Under Pressure. Since ordinary rocks are brittle, 
they will fracture and crush under high pressures in the laboratory unless 
special precautions are taken. When a rock specimen is confined on all 
sides, and subjected to very intense compression, its size decreases 
slightly; in other words, rocks are somewhat compressible. If a marble 
core is fitted into a cylindrical opening in a strong steel jacket, and 
subjected to enormous pressure by means of steel pistons, in time the 
walls of the steel jacket are bulged outward. By cutting away the 
jacket it is found that the marble core has been shortened and thickened, 
but not crushed (Fig. 262). Evidently it has flowed slowly, as if it were a 
plastic substance. Since rocks deep in the Earth are confined under high 
pressure, it is argued that under certain conditions these rocks are de- 
formed by plastic flow. Strength and rigidity are only relative terms. 


Density Distribution. It is certain that the outer part of the Earth 
consists of much lighter material than the average. Several assumptions 
might be made, however, as to the arrangement of light and heavy sub- 
stances between the surface and the center. For example, it is possible 
that light rocks, like granite, form a shell a few miles or tens of miles 
deep, and below this shell the Earth is composed of very heavy rock with 
a uniform density of approximately 6. Again, it may be assumed that 
from the outer zone of low density there is a gradual and progressive 
increase to a density of 9 or 10 at the center. By either of these arrange- 
ments the average density of 5.52 might result. Besides these two sug- 
gestions a number of other assumptions as to distribution of density 
might be made, all of them consistent with the known average density. 

Fortunately, there are other checks to guide us in attacking the prob- 
lem. Any assumed distribution of the density must harmonize with the 
mathematical and mechanical knowledge to which reference has been 
made above. The small degree of flattening at the poles of the Earth 
suggests that a large percentage of the mass is concentrated in the central 
portion; and the same suggestion is evident from study of the preces- 
sional effect. These considerations, therefore, favor the assumption 
that density increases slowly downward in the Earth, and that the cen- 
tral core is composed of very heavy material. 

Once this point is reached in the inquiry, another problem presents 
itself. Is the increase of density toward the center of the Earth due 


wholly to compression under enormous weight, or is there a concentration 
toward the center of metals that normally are heavy? Some students of 
the problem have argued that compression of ordinary rock material is a 
sufficient explanation. At a depth of one mile, each square foot of rock 
bears a weight of 450 tons; and with each additional mile the pressure is 
increased by more than this amount, as the material grows progressively 
denser. Near the center, pressures amount to more than 3,000,000 tons 
per square foot. Without question such intense compression has an 
effect in compacting the material. Since no pressures that are possible 
in laboratory experiments can even remotely simulate conditions deep 
within the Earth, we cannot state positively how much compacting can 
result. However, from various lines of evidence and reasoning it ap- 
pears unlikely that ordinary rocks can be compressed sufficiently to give 
the high average density of the Earth. Therefore many scientists favor 
the view that the core of the Earth is in large part metallic. 

Which of the metals is most likely to exist in such abundance in the 
Earth? The most satisfactory answer to this question comes from 
consideration of the material that reaches us from outer space. The 
majority of known meteorites are composed of nickel and iron, and the 
others consist of dark-colored, heavy rock. Since these bodies probably 
were derived originally from the sun, along with the material in the 
planets, it is argued that they suggest the composition of the Earth. 
Adopting this argument, and using all available information, certain 
scientists have postulated the following arrangement in the Earth: 

(1) An outer layer about 35 miles thick, in which the material changes 
gradually from granite to dark rock somewhat heavier than gabbro; (2) 
a zone extending to a depth of nearly 1000 miles, consisting of peridotite, 
with a density ranging up to more than 4; (3) a zone reaching to a depth 
of nearly 2000 miles below the surface, in which peridotite is gradually 
replaced by iron or nickel-iron, with a density of 9.5 at the base of the 
zone; and (4) a central core of nickel-iron, with density about 10. A 
part of the increasing density with depth is attributed to compression 
under load, but a larger part to materials that normally are heavy 
(Fig. 263). 

It should be kept in mind that this suggestion is merely a hypothesis, 
no part of which is subject to proof at present. It is only an attempt to 
give a picture that is consistent with all known facts. 

Temperatures at Depth. The average change in temperature for a 
given unit distance is known as the temperature gradient. If the gradient 
determined in mines and bore holes should continue downward un- 
changed, the temperature of the center would exceed 350,000 F. Aside 
from the fact that this figure seems inconceivably high, several consider- 



ations make It improbable that temperatures in the interior are so ex- 
cessive. The length of the deepest opening used in estimating the gradi- 
ent is less than T V?r tne Earth's radius; and the value obtained for this 
thin skin cannot be accepted with any confidence for the whole body. 
Rocks such as granite are extremely poor conductors of heat. Therefore 
if the temperature at a depth of several miles should be high, say 1000 
or 1500, heat would flow out very slowly, and the change in temperature 
for each 100 feet would be considerable. It is altogether likely that the 
heat conductivity improves with depth, both because high pressure 
compacts the rocks and because metallic substances, which are notably 

Fig. 263. To suggest the concentration of metallic substances toward the center of 
the Earth, (&), outer zone of granite and basalt; (6), zone of peridotite; (c) peridotite 
mixed with nickel-iron. The black central portion represents a core of nickel-iron. (Adams 
and Williamson.) 

good conductors, probably grow more important in the deeper zones. 
Thus, whatever the temperature of the deep interior, it is likely to be 
distributed more uniformly than in the outer zone. To express this 
conception another way, let us imagine that the core of the Earth is very- 
hot. If the structure and composition are as represented in Fig. 263, the 
heat will flow out rapidly through the zones made of metal and of , heavy 
rock, producing a nearly even temperature; but on reaching the outer 
zones, made of granite and other poor conductors, the flow of heat will be 
checked, and the temperature will fall rapidly in approaching the surface. 
Therefore the gradient in the shallow zone probably is larger than for 
any other part of the globe, and cannot be used to calculate the tempera- 
tures at great depth. 


It is even probable that much of the heat conducted to the Earth's 
surface and lost by radiation does not come from great depth, but is 
generated in the shallow zone. The presence in granite and other known 
rocks of uranium and other radioactive elements has been mentioned. 
These elements disintegrate slowly but continuously, with evolution of 
heat. The possible significance of this process in connection with igneous 
activity has been discussed previously. 

Whatever may be the true value for the average temperature gradient, 
many lines of evidence suggest that temperatures in the interior are high. 
Probably they are above the melting points for the materials under sur- 
face conditions; but as earthquake waves testify that no considerable 
part of the globe is molten in the outer 2000 miles of its radius, it is certain 
that heat is not great enough in this portion to exceed the "critical tem- 
perature" under the great pressures that prevail. Locally this control 
by pressure may be overcome in one of two ways: by actual decrease of 
the pressure, or by unusual concentration of heat. According to our 
conceptions, pressure may be relieved only at shallow or moderate depth, 
by local arching up of the superficial rocks or by deep fracturing. The 
most conceivable reason for exceptionally high temperature in the rocks 
is the presence of radioactive substances in unusual amount. As our 
reasoning confines these substances to the outer part of the Earth, it 
appears that conditions for actual liquefaction of rocks are favorable 
only at shallow or moderate depths. From local pockets of magira 
generated by any cause the surplus heat is dissipated in time, either by 
volcanic activity or through movement of the magma into higher and 
colder rocks. Thus the temperature finally is reduced below the critical 
point, and the solid condition is resumed. 

The condition of matter in the central core is still a subject for varied 
speculation. From the testimony of earthquake waves we reason that 
the core is not an elastic solid. Some scientists argue that the interior 
heat is sufficient to keep the metallic core fused in spite of the enormous 
pressures at that depth. 

Isostasy, or Equittbrium in the Crust. The Earth does not have the 
form of a cube or a pyramid or other angular figure, for a good mechanical 
reason. Rock is not indefinitely strong. Under great strain it adjusts 
itself to a condition of equilibrium, as does any other material. If the 
Earth could by any means be forced momentarily into a greatly distorted 
shape, the laws of Nature would bring about restoration to a figure of 
equilibrium. This figure would be essentially a sphere if there were no 
rapid rotation; but the spinning on an axis makes the figure an oblate 
spheroid, with the degree of flattening determined by the rate of spinning. 

How high may a mountain mass rise above the general surface, or how 



far may an ocean "deep" lie below it? Certainly not to unlimited height 
or depth, for then the natural law would be violated. The strength of 
rock is sufficient to bear considerable strain, but there is a definite limit, 
as may be demonstrated by laboratory experiment. Is this limit ever 
exceeded in the outer part of the Earth? In geology we see the evidence 
that great masses of material are shifted from one part of the surface to 
another. Rocks are folded and crowded together by horizontal move- 
ments in mountain zones; continents and mountains are eroded, and the 
debris is piled along the continental margins,- more than once vast quan- 
tities of water have been removed from the oceans and heaped upon the 
continents in the form of thick ice sheets. These transfers of material 
tend to change the form of the spheroid, and undoubtedly set up great 
strains. Is there some mechanism for adjusting these strains? 

This view of the subject is altogether deductive. There is also a more 
practical avenue of approach, through geodetic measurements and geo- 

Fig. 264. Effect of deficient density in a mountain mass. OA represents the true 
vertical; OC the actual position taken by a plumb line; OB the calculated position of the 
plumb line assuming rock of uniform density under mountains and plain. Angles and 
vertical scale much exaggerated. 

logic evidence. The geodesists use two important methods of investi- 
gation. A plumb line suspended on a plain near a great mountain front 
is attracted laterally by the upstanding mass, and therefore does not 
point exactly toward the center of the earth (Fig. 264). Knowing the 
volume of the mountain unit, and assuming that the rocks in and be- 
neath the mountains have the average density of normal rocks at the 
surface, we can compute accurately the amount of lateral attraction to 
be expected. In actual experiments the deviation of the plumb line from 
the vertical is only a small fraction of the expected value; and therefore 
we reason that the rocks in or under the range are abnormally light. 
This conclusion is checked by precise measurements of gravity on the 
high and low areas, by the use of a delicate pendulum. The rate at 
which the pendulum vibrates is governed closely by the intensity of 
gravity, which varies at different localities with latitude, height above 
sea level, and other known factors. By considering these factors, 


geodesists can calculate closely what the value of gravity at any station 
should be; but for mountain stations it is found in general that the 
calculated values are too high. Again the result is explained by assum- 
ing that the excess volume of material represented by the mountains 
is offset by deficient density beneath. It is inferred, therefore, that 
high areas such as continents or mountain chains are more or less 
balanced against low areas such as ocean basins or low plains. The 
term isostasy (from the Greek meaning " equal standing ") is used for 
this supposed condition. 

A simple illustration of the principles involved in isostasy is given in 
Fig. 265. The different metals vary considerably in density. Therefore 
if blocks are taken with the same weight and cross section, blocks of the 
light metals are considerably longer than those of heavy metals. All of 
these float in mercury, which is 13.6 times as dense as water. As the 
blocks have equal weight, they sink to the same depth, leaving the top 

3 siiver 


Lead bjoS? 







Fig. 265. Diagram to illustrate an irregular upper surface on floating blocks of differ- 
ing density. The blocks are equal in cross section and in weight, and therefore sink to 
equal depth. This is an ideal illustration of one isostatie theory, which assumes that differ- 
ences in altitude between major features of the Earth's surface are compensated by differ- 
ences in density down to a certain level, below which the density is essentially uniform. 

surfaces at irregular heights. The longer blocks might be taken to 
represent mountains and plateaus; the shorter, low plains or basins. It 
is not to be understood, of course, that the Earth's crust is divided into 
definite blocks, or that there is a liquid substratum at some depth. The 
illustration is highly artificial, and any attempt to press the comparison 
closely will result in misconception. It is intended merely to emphasize 
the principle of mass balance. 

Figure 265 may suggest that great differences in topography reflect 
large variations in kinds of rock at the surface. From geologic evidence, 
however, it appears that the commonest rocks everywhere in the con- 
tinents are granites. Therefore it is possible that mountains represent 
merely a local thickening of the granitic crust. This conception is illus- 
trated in Fig. 266, in which all the blocks are of copper, but of different 
lengths. Irregularities at the surface are reflected by similar inequali- 
ties reaching downward. This general arrangement applied to the Earth 
would satisfy the principles of isostasy as well as the conception in Fig. 
265, and would accord better with geologic evidence. 



Suppose now that a portion is cut from the top of one block (in either 
figure) and placed on another. Equilibrium is disturbed, and adjust- 
ment takes place by sinking of the loaded unit and rising of the other. 
In a liquid this adjustment is immediate and perfect. How can the 
principle apply in the Earth? We are satisfied that no continuous zone 
of liquid rock exists; but it is known that solid rock behaves as a plastic 
substance under high pressure. The laboratory proof has been men- 
tioned, and circumstantial evidence of rock flowage is seen in metamor- 
phic rocks exposed by deep erosion. It is inferred, then, that any over- 
loaded part of the crust sinks, displacing the deep rocks by plastic flow 
and thus causing lighter parts of the crust to rise. The operation is 



~~-f z 



------ --By- 

Fig. 266. Copper blocks, equal in cross section but unequal in length, float in mercury. 
They sink to unequal depth, and also rise to unequal height. It can be shown that if the 
mercury directly under each block down to the line A-B is included, all the blocks are^of 
equal weight. According to this conception of isostasy, a continent consists of a granite 
shell, essentially uniform in density but with variations in thickness. The thinner por- 
tions form low plains, whereas thicker parts project upward as plateaus and mountains. 

much less perfect than in a liquid and of course requires a much longer 

This assumed mechanism for preserving balance in the crust is illus- 
trated by Fig. 267. A mountain mass is eroded deeply, and the d6bris is 
transported across an adjacent low plain, to be deposited in an ocean 
basin. For a time the crust can bear the strain, but eventually the 
mountain segment becomes abnormally light through loss of mass; the 
surrounding crust, and especially the part loaded by the sediments, 
forces deep-seated rocks to move laterally and buoy up the lightened 
segment. At first thought it may appear that this mechanism would 
make it impossible for erosion ever to wear highlands to a low level, as 
there would be constant rejuvenation. However, as the deep rocks are 
denser than those near the surface the amount of uplift cannot equal the 
thickness of lighter rock removed; and therefore mountains can finally 
be brought low, though only by long-continued erosion. 

A striking confirmation of the isostatic theory is furnished by the 
areas in North America and Europe that were covered by great ice 
sheets during Pleistocene time. As the glacial ice melted the sea invaded 
much of the glaciated area, but was excluded by later uplift. Bones of 
whales and dolphins, and other evidences of this late submergence, are 
found in the region of Lake Champlain and Montreal, 600 to 700 feet 


above present sea level; and similar evidence Is found in Scandinavia. 
The plain inference is that the glacial ice was an overload which depressed 
the land. After the load was removed some time was required for res- 
toration of balance by slow plastic flow in the rocks at depth. Although 
it is thousands of years since the ice disappeared, perhaps the adjustment 
is not yet complete, as parts of Scandinavia are still rising at the rate of 
2 or 3 feet in a century. Strong tilting of the Great Lakes basins in 

Direction of movement of eroded material 

Probable direction of movement of material to ""*- 

maintain equal weights of earth blocks 

Fig. 267. Diagram to explain the supposed mechanism to restore balance when load 
is shifted on the surface by erosion. The eroded area, at the left, continually loses weight, 
which is added to the low area at the right. In time this loaded part of the crust sinks, and 
forces deep-seated rock to move horizontally by slow plastic flowage. This material moves 
underneath the eroded "block," which rises to a position of proper balance. The plains 
area, across which eroded sediments are transported, suffers neither depression nor uplift. 
The representation is highly artificial, as there are no definite, freely moving "blocks" 
in the lithosphere. (Bowie.) 

North America is ascribed to the postglacial uplift, which seems to have 
been greatest in southern and eastern Canada, where presumably the 
glacial load was greatest. 

The general fact of isostasy appears to be established; but many un- 
certainties are connected with the subject. We do not know the depth 
at which plastic flow occurs during adjustment, though from several 
lines of reasoning it is argued that this depth is only a few tens of miles. 
The size of a load necessary to start the mechanism of adjustment is not 
known. These and other problems may be solved by continued study. 


The problems connected with the Earth's interior are fascinating but 
difficult. Their solution requires the cooperation of geology, physics, 
mathematics, and other branches of science. Perhaps some of them are 


quite insoluble; but a promising beginning has been made. It is well to 
keep in mind that some of the present views and conclusions are tenta- 
tive, and may be changed by continued investigation. Speculations in 
this field are numerous, and these should be kept distinct from legitimate 
inference and proved fact. 


1. OBT Mobile Ea,rth; by R. A. Daly. Scribner's, New York, 1926. Chapter 
III (pp. 90-127) discusses evidence bearing on "The Earth's Interior" and outlines 
the author's views on the subject. 

2. The Composition of the Earth's Interior; by L. H. Adams and E. D. William- 
son. Smithsonian Report for 1923, pp. 241-260. 

A brief statement of the evidence from various sources, followed by the authors' 
suggestions on the composition of the Earth at various depths. 

3. On Some of the Greater Problems of Physical Geology; by Clarence E. Dutton. 
Bulletin Phil. Soc. Wash., Vol. 11, 1889, pp. 51-64. Reprinted in Jour. Wash. Acad. 
Sci., Vol. 15, 1925, pp. 359-369. 

A classic paper, admirably written, in which the term isostasy was first proposed 
and defined. 


Mountains are of great importance in geology, as they furnish, a large 
part of the information on which the science is based. Dynamic proc- 
esses, such as stream erosion and glaciation, are especially vigorous and 
their effects strikingly evident in high ranges. The uplifting of rock 
masses to great heights has resulted in dissection to unusual depth; and 
consequently a mountain region affords excellent opportunity not only 
for descriptive study of rock formations, but also for deciphering the 
history they record. But if the mountains give aid in solving many 
problems relating to the Earth, they also present mysteries in themselves. 
Why have sea floors of remote periods become the lofty highlands of to- 
day? What generates the enormous forces that bend, break, and mash 
the rocks in mountain zones? These questions still await satisfactory 
answers; but the architectural features of great ranges at least offer hints 
as to their origin, and are worthy of study for their own sake. 


An isolated high mass that rises above comparatively low surroundings 
is described simply as a mountain or a peak: Examples are Stone Moun- 
tain in Georgia, Mount Monadnock in New Hampshire, and Mount 
Etna in Sicily. More commonly mountain masses do not stand alone, 
but are parts of distinct units that vary in size and plan, from small 
irregular groups, like the La Sal Mountains of Utah and the Black Hills 
of South Dakota, to the enormous belt that extends more or less regu- 
larly from Gibraltar eastward to the East Indies. Descriptions of the 
larger units or their parts employ somewhat loosely the terms range, 
system, and chain. As it is desirable to use descriptive terms with a 
definite meaning, the usage proposed many years ago by J. D. Dana will 
be followed generally. 

A mountain range is either a single large, complex ridge, or a series of 
neighboring parallel ridges that form a more or less continuous and 
compact unit. Excellent types are the Sierra Nevada in eastern Cali- 
fornia and the Front Range of Colorado. A group of ranges that are 
obviously similar in their general form, structure, and alignment, and 
presumably owe their origin to the same causes, constitutes a mountain 

381 * 


system. Thus the Southern Rocky Mountain system, extending from 
Wyoming through Colorado into New Mexico, is made up of the Sangre 
de Cristo, Front, Sawatch, and other great ranges formed in the same 
geologic period. A different designation is needed for a mountain belt 
such as the Appalachians, which includes a number of groups and ranges 
diverse in plan, structure, and geologic date. Accordingly all of the 
mountains in the broad belt extending from Alabama to Nova Scotia and 
Newfoundland are grouped together as the Appalachian chain. The 
Rocky Mountain chain includes both the Southern and the Northern 
Rocky Mountain systems, constituting a great unit that extends from 
near the Mexican boundary through the United States and western 

But a still more comprehensive term is needed in referring to a series 
of chains, systems, and ranges that make a more or less compact belt of 
large extent. Following the famous traveler Humboldt, a Spanish word 
has been borrowed for this purpose. All of the mountain units in western 
North America, from the eastern border of the Rocky Mountains to the 
Pacific coast, are known as the North American cordillera. Similarly 
the entire broad mountain belt that extends almost continuously from 
Alaska to Cape Horn is known as the American cordilleras. However, 
the same term has not been adopted universally for the major mountain 
belts of the Earth. In the literature the great mountain unit of southern 
Europe and Asia is designated variously, as the Mediterranean (or 
Eurasiatic) chains, zone, or belt. 


Mountains owe their origin to different agencies, and differences in the 
structure and the plan of mountain units are due largely to this fact. 
Ultimate causes of mountain building are in large part obscure; but re- 
gardless of this fact the principal agencies involved may be stated as 
differential erosion, volcanic activity, and movements of the crust. 
Commonly two or more processes combine to produce complex results. 
However, the discussion will be clarified by a classification that recognizes 
the dominant agencies. 


High plateaus suffer differential erosion, and during a late stage in the 
process some of the more favored residuals may have sufficient height, 
in relation to their surroundings, to be called mountains. Some of the 
larger buttes in western United States are examples. The Catskill 
Mountains in New York represent remnants of an extensive high plateau, 



the greater part of which has been removed by stream erosion. An early 
stage in a similar development may be seen in the Grand Canyon of the 
Colorado, where a number of pyramidal erosion remnants rise from the 
depths of the chasm (Fig. 268). These are dwarfed by their surround- 
ings; but if they could be placed on a plain they would rise to mountain 
heights. It is not difficult to imagine that in a later geologic epoch the 
present youthful plateau will have been dissected thoroughly by the 

Fig. 268. Vishnu's Temple, Grand Canyon, Arizona. Mountain-like remnants, left 
during dissection of Colorado Plateau by Colorado River and tributaries. Part of the 
level, undissected plateau, many miles distant, appears in background. (U. S. Geo- 
logical Survey.) 

Colorado and its tributaries, and will be represented only by scattered 
residual mountains separated by plains and valleys. 

Plateau blocks from which residual mountains are derived may consist 
of horizontal rock formations, of homogeneous crystalline rocks, or of 
folded and faulted strata that were peneplaned by previous erosion. 
Obviously the resulting mountain forms will be influenced by the original 
structure. Residual mountains are also known as "mountains of ero- 
sion." In a strict classification, however, it must be considered that 
deep differential erosion is made possible only by great uplift, and there- 
fore movement of the crust is an important factor in producing residual 
mountains. As plateaus, in their initial form, are entirely distinct from 
mountains, a rigid classification might insist that the erosion residuals be 


called "plateau remnants" and not mountains. However, this would be 
an academic distinction without practical value. 


Some of the loftiest peaks in the world, such as Chimborazo (20,517 ft.) 
and Aconcagua (23,393 ft.) in the Andes, and Kilimanjaro (19,321 ft.) in 
Africa, have been built directly by volcanic action. Many such peaks, 
however, have their bases on high plateaus, and therefore their true 
height is much less than the altitudes above sea level indicate. A large 
number of volcanic islands are great volcanic piles, and some of them 
appear to be seated directly on the deep ocean bottoms. Thus the island 
of Hawaii, measured from the Pacific floor to the highest peaks, has a 
total height of about 30,000 feet; and the entire mass, so far as can be 
judged, consists of extrusive basaltic rocks. 

Volcanic peaks commonly are superimposed on mountains that re- 
sulted from other causes. All mountain masses formed directly by 
igneous extrusion are sometimes called "mountains of accumulation." 
In areas of widespread volcanism the volcanic materials build up ex- 
tensive plateaus, such as the Absaroka Plateau east of Yellowstone Park, 
and the great Columbia Plateau of Washington and Oregon. From units 
of this kind residual mountains are produced in time by differential 


The mountain types considered above are important geologically, as 
they form groups of considerable size in the world today and have had 
wide distribution during past geologic periods. Differential erosion is 
one of the major factors in producing mountain relief. However, this 
factor could not operate unless large areas of the continents were raised 
high above sea level. Furthermore in most of the dominating mountain 
units the relief has been conditioned, either directly or indirectly, by 
localized movements which have caused more or less severe disturbance 
of the rocks. In some of the young mountain systems much of the local 
relief was caused directly by these movements; in older belts, which have 
experienced deep denudation and perhaps more than one rejuvenation by 
regional upwarping, the original disturbance of the rocks may be im- 
portant chiefly in guiding erosion. But even if the present mountain 
relief is due chiefly to differential erosion, any characteristic structure 
produced by crustal movement has large importance in classifying moun- 
tain units. According to the nature of the movements, as indicated in 
the resulting structure and form, mountains of this general type may be 


divided into four classes. (1) Dislocation or faulting on a large scale 
results in relative uplift of rock masses, with or without tilting. Ranges 
whose structure is produced chiefly by this process are fault mountains. 
(2) Some vertical movements result in arching of the rocks into a general 
domal form, either nearly circular or somewhat elliptical in plan. Dome 
mountains result from this process. (3) More commonly the forces 
deforming the crust produce large plications, or parallel anticlines and 
synclines, giving rise to the structure of fold mountains. (4) In most 
of the great mountain belts we see the combined effects of two or more 
types of movement, particularly folding and faulting, with complications 
produced by igneous intrusion. The resulting mountains are of the 
complex type, although locally they may be classified according to the 
process that has played a dominant role. 

Excellent examples of each mountain type exist; but Nature is com- 
plex, and consequently various combinations of the different types are 
most cqmmpn. ' The influence of erosion in varying degree is evident in 
all mcMi^fes^'^egaTdtess of type. 

J - :^ Fault Mountains 

Assume that a' system of intersecting fractures, reaching to great 
depth, divide part of the Earth's crust into blocks or masses of very large 
dimensions. Mountains may be produced directly by movements of 
several kinds, (a) If the region is initially a high plateau, some of the 
blocks may be depressed several thousand feet, leaving other blocks in 
their original positions, to form mountain ranges. (&) Regardless of 
original altitudes, some of the blocks may be elevated to mountain 
heights, by forces acting largely in the vertical direction, leaving adja- 
cent blocks relatively depressed, (c) All of the blocks may move down- 
ward or upward, but differentially, so that in the end some stand much 
higher than others, (d) Each of several blocks may be tilted or rotated, 
one edge being elevated and the opposite edge depressed. The faults on 
which movement occurs may be either normal or reverse. Regardless 
of the nature of movement, a series of neighboring ranges obviously 
due to faulting may be called either fault mountains or block mountains 
(Fig. 269). 

Fault blocks as we actually see them have been more or less modified by 
erosion. Debris worn from the high masses tends to bury those at low 
elevation. In time this combination of erosion and deposition may 
nearly or quite obliterate the mountain relief, especially in a region of 
interior drainage where all the debris is retained. At a later date, as a 
result of broad regional uplift perhaps accompanied by change of climate, 



the original pattern of ranges and intermontane troughs may be etched 
out by vigorous erosion. Obviously these resurrected ranges are the 
direct result of differential erosion, and strictly they are residual moun- 
tains. However, it is recognized that the original faulting was a strong 
factor in guiding denudation and continues to be reflected in the moun- 
tain forms. Accordingly ,the original structure is given large emphasis, 

Fig. 269. An example of fault mountains. A great fault zone, somewhat irregular, 
marks the base of the range. Down/thrown block, ,at left, covered with lake deposits. 
Searp 2000 feet high. Bluejoint Rim, Oregon. (Airplane view by H. E. Fuller.) 

and it is common to use the designation fault mountains or block moun- 
tains even for units that have been greatly modified by erosion. 

The Sierra Nevada of California is a tilted crust block 400 miles in 
length and approximately 100 miles wide. Its eastern edge has been 
uplifted two miles or more, to form an abrupt eastward-facing scarp. 
Roads from the east ascend this precipitous front with difficulty; but 
west of the crest they descend on a long, gentle slope the tilted upper 
surface of the block. In the Great Valley of California, sediments 
thousands of feet deep have accumulated on the depressed portion of 
the rotated mass. A great series of fault mountains lies in the Great 
Basin of Nevada and neighboring states. Ranges in that region are so 
similar in character that they are known collectively as the Basin and 


Range System. The great dislocations responsible for these ranges and 
for the Sierra Nevada do not represent the first disturbance of the 
region. In earlier geologic periods the thick sedimentary rocks in Nevada 
and eastern California were folded and crushed, and great igneous 
masses were intruded into them. Mountains that existed soon after 
those ancient events disappeared long ago, and in fairly recent time the 
old deformed crust has been broken by great faults, to form the present 
generation of mountains. Some of the ranges are still growing; for 
within historic time movements have occurred on several of the faults, 
giving rise to violent earthquakes. As a large part of the Basin and 
Range region has no drainage to the sea, the mountains are partly buried 
by accumulations of their own debris. 

Fault mountains in various stages of destruction are found in parts of 
eastern and northern Africa, in Arabia, and in central Asia. The Trias- 
sic sandstones of Connecticut and Massachusetts are broken by great 
faults, and the resulting blocks have a strong tilt eastward; but the 
mountains that were formed by these dislocations have disappeared 
through erosion. Similar structural relics of ancient fault mountains, 
representing various geologic periods, are widely distributed in all 

Dome Mountains 

Mountains whose structure reflects crustal uplift of distinctly domal 
character may be classed together, regardless of size or the exact cause of 
the uplift. The simplest and best understood are laccolithic domes, 
made by the bowing up of strata above thick, lens-shaped intrusions of 
liquid rock. Ordinarily a dome of this kind that is high enough to be 
called a mountain has lost more or less of its original cover through 
erosion; and not uncommonly the resistant igneous mass, almost com- 
pletely denuded, stands within circular or elliptical ridges formed by the 
upturned edges of the more resistant strata. An excellent example is 
Bear Butte, one of many laccolithic mountains in the vicinity of the 
Black Hills (Fig. 190). The Henry Mountains of Utah, classic examples 
of the type, are a large group of laccoliths in various stages of denuda- 
tion. But not all mountains of this kind have ideally simple structure. 
The intruding magma not uncommonly ruptures the covering strata and 
lifts them irregularly, as in the Moccasin Mountains of Montana. A 
compound laccolith, such as the Mount Holmes mass in western Colorado, 
presents, after some erosion, a confused arrangement of the igneous rock 
and the intruded strata. 

Dome mountains on a larger scale are illustrated by the Black Hills of 
South Dakota. In a casual journey through these mountains the 


traveler may gain an impression of disordered arrangement; but a good 
map or a block diagram of the entire unit reveals a beautiful symmetry 
of structure, involving an elliptical area about 100 miles in length by 50 
in width (Fig. 270) . In the middle and eastern portions the uplifted sedi- 
mentary formations have been stripped away, and there the crystalline 
basement rocks have been carved into ridges and peaks, including 
Harney Peak (7242 feet above sea level), the highest point east of the 
Rockies. Around the flanks of the uplift the upturned edges of sedi- 
mentary formations, yielding to erosion at different rates, make alter- 

Fig. 270. .Diagram of the Black Hills uplift, South Dakota. View looking north, 
along the longer axis of the elliptical dome. The wide valley encircling the dome is 
commonly known as "The Race Track." (Newton, U. S. Geog. and Geol. Survey.) 

nating high and low belts which encircle the uplands. Some of the heavy 
limestone members still cover the western half of the uplift, forming the 
"limestone plateau." A careful profile and section, in which the eroded 
formations are restored to their original positions, shows that the dome is 
steeper on the eastern than on the western flank, and that the greatest 
erosion has occurred in the area of maximum uplift. The restored 
section indicates that the top of the dome was elevated 9000 feet, although 
the present height above the surrounding plains is less than 4000 feet. 

In contrast with a laccolith, the crystalline rocks exposed at the core 
of the Black Hills were not intruded as magmas, but are much older than 
the domed sedimentary formations. Therefore the force by which the 
uplift came about was applied much deeper than any exposed part of the 


crust. If the cause of uplift was the rise of igneous material, the molten 
mass was very large, quite symmetrical in form, and very deep-seated. 
The connection of some igneous activity with the movement is indicated 
by numerous laccoliths, which form small satellitic domes in the northern 
part of the Hills. Whatever may have been the part performed by 
igneous magmas in causing the main uplift, it is very probable that the 
horizontal pressure by which the Rocky Mountains were folded was an 
important factor also in shaping the Black Hills dome. The long axis of 
the Hills is parallel to the Rocky Mountain front, and the eastern side 
of the dome is especially steep, as if there had been strong thrust from the 
west. According to this view the bulging up of the Black Hills was 
merely a local incident in forming the Rocky Mountain structure; and 
the laccoliths superimposed on the larger dome represent still smaller 
incidents in the general process. 

It is evident, on reflection, that the Black Hills uplift did not become 
an actual mountain group without the work of erosion. If the youngest 
strata involved in the movement still extended unbroken across the 
summit, forming a broad, smooth dome, the area would be a small 
plateau rather than a mountain. It is essential, therefore, to keep in 
mind the limitations of any scheme of classifying mountains. The part 
played by erosion is of great importance in connection with every other 
process. However, the crustal deformation which controls the relief 
forms is important enough to merit recognition in a classification. The 
limit of size separating dome mountains from plateaus must be somewhat 
arbitrary; but as the Black Hills uplift is a definite unit of moderate size, 
with strongly defined boundaries, it seems proper to emphasize the 
structural form, provided the various steps in fashioning the highlands 
are kept clearly in mind. 

If the Black Hills should be worn down to a peneplain by prolonged 
denudation, and a later warping movement should reelevate the Great 
Plains region several thousand feet, subsequent erosion would be guided 
by the old structure and a group of residual mountains similar to the 
present Hills would be produced eventually. 

Fold and Complex Mountains 

Mountains in which the rocks are strongly folded and broken are 
commonly described according to their internal structure, regardless of 
the later chapters in their history. Some old mountain units may be 
strictly remnants of erosion, and therefore residual mountains. How- 
ever, if they give evidence that a certain type of crustal deformation 
attended their early development, these structural characteristics are 


used as the basis of their classification. For example, the Appalachians 
have had a long and varied career. The original chain suffered erosion 
through long ages, and almost or quite disappeared; and the present Ap- 
palachian ridges have been etched out after a later upwarping of the 
area containing the old mountain "roots." Nevertheless, remnants of 
the original structure are clearly visible, and are recognized as an im- 
portant feature of the mountain zone. It is well to keep this example in 
mind throughout the discussion that follows. Deformed rocks are 
characteristic of all great mountain zones, and the development of this 
deformed structure will be given prominent attention in outlining moun- 
tain history; but it is not to be taken for granted that the deformation 
gave rise to the present mountain elevations. Mountain structure and 
mountain elevation may not have any direct relation to each other. 
Nevertheless the structure continues as a dominant factor in determining 
relief because it guides differential erosion. 

All the great mountain chains of the Earth include folded sedimentary 
rocks as a conspicuous part of their structure. These chains, therefore, 
are sometimes classed together as fold mountains, although faulting, 
igneous intrusion, and other important processes besides simple folding 
have played some role in their origin. Strictly speaking, every great 
system is more or less complex in its structure; but certain mountain units 
exhibit fairly regular plication of rock formations as their outstanding 
structural characteristic. The Jura Mountains in Switzerland and parts 
of the Appalachian Mountains in North America are excellent examples. 
The Rocky Mountains and the Alps are characterized by enormous 
thrust faults in addition to folds, and consequently are illustrations of 
complex units. However some parts of the Appalachians also are com- 
plicated by thrust faulting; and as there are all stages of gradation be- 
tween the simpler sections of this chain and the almost incredible com- 
plexity of the Alps, it is clear that fold and complex mountains cannot 
be separated as sharply contrasted structural types. Therefore it is 
desirable to include mountains of these two classes in a unit discussion; 
although the treatment logically emphasizes the simpler folding first, and 
then proceeds to more complex processes and results. 

Considering the Earth as a whole, the finest exhibitions of geologic 
phenomena are furnished by the mountains with folded and complex 
structure. There are to be found the upturned and dissected strata 
whose kinds, thickness, included fossils, and structure furnish the most 
effective key to past events. Commonly the making of such mountains 
has been accompanied by igneous activity, and the sections now exposed 
reveal both intrusive and extrusive masses of various types. Many of 
the most important ore deposits occur in these zones of disturbance, both 


recent and ancient. Some of the youngest complex ranges are the thea- 
ters in which many agents of erosion, as well as crustal movements and 
volcanism, play their most active roles at the present time. For many 
reasons, therefore, the great mountain belts merit special consideration. 

General Characteristics of Fold and Complex Mountains. From 
examination of a globe or a world map it is apparent that the prominent 
mountain belts are elongated generally parallel to the continental mar- 
gins. This relation is especially striking in the American Cordilleras, 
the Appalachians, the Scandinavian chain, and the great Eurasiatic 
mountain zone. Each major belt is composed of numerous ranges dis- 
posed somewhat irregularly but with the same general orientation. Some 
of the ranges are nearly straight in plan; but many are strongly curved 
into the form of great bows or arcs. The Alps, Carpathians, and 
Himalayas are striking examples of this arcuate type. 

Generally the exposed portion of each range is made up in part or 
wholly of distorted sedimentary formations. Commonly these strata, 
now on the flanks or even on the highest summits of the ranges, represent 
deposits in former seas or on deltas and in marshes bordering the sea. 
Owing to the strong folding and faulting of these strata, followed by 
planation and dissection through erosion, the full thickness of the sedi- 
mentary cover is exposed in many places. In some mountain belts 
these thicknesses are astonishing; 4, 5, or even 6 miles are by no means 
exceptional values, and in some mountain areas the total sedimentary 
sections exceed 40,000 feet. It will occur to some readers that similar 
thicknesses may be common also outside of mountain zones, but are not 
known because conditions favorable for their revelation do not exist. 
However, natural exposures and well records indicate clearly that sedi- 
mentary formations grow conspicuously thinner away from a folded 
mountain belt. Thus the strata in the Appalachians average 20,000 
feet or more in thickness along the central axis of the folded tract; but at 
no great distance to the west the thickness is less than 10,000 feet, and 
in the Mississippi Valley it is only 4000 to 5000 feet. On the east side of 
the Appalachians the sedimentary strata do not exist, and it will be 
shown presently that the deposits never extended far eastward from their 
present limit. Therefore the excessively thick sediments occupy a long 
and relatively narrow belt that corresponds closely to the axis of the 
folded chain. This general relationship exists also in the Rockies, the 
Andes, the Mediterranean ranges, and other great mountain systems. 
It is a natural conclusion that the accumulation of abnormally thick 
sediments had a significant connection with the development of each of 
these mountain units. 

The typical history of ranges with fold structure falls into three general 


divisions: the preliminary stages in which certain processes prepare the 
place and some of the material for the future mountains; the period of 
crustal movements, in which the folds and related structural features are 
produced and the initial uplift occurs; and subsequent stages, during 
which the mountains experience various modifications through erosion 
and repeated vertical movements. This subdivision of the history is 
useful, if it is understood that the periods are not sharply separated, and 
that some of the processes involved operate simultaneously. Thus some 
incidents of folding and thrusting occur long before the accumulation of 
sediments is complete; and inevitably the modifying influence of erosion 
dates from an early stage of uplift, while the orogenic or mountain- 
making processes are active. The growth and decline of every range is 
the result of a slow, complicated interplay of forces that act either con- 
tinuously or recurrently during a very long geologic interval. Recog- 
nition of general stages in the history serves to clarify a brief discussion; 
but these stages overlap and merge irregularly into one another, so that 
the whole sequence is typified by the cross profile of a mountain chain 
itself, in that it rises gradually and with interruptions to a culmination 
and declines in the same way. 

Preliminary Stages: Development of Geosynclines. Sedimentary 
strata in the great ranges consist of conglomerates and sandstones min- 
gled with shales and limestones. It is clear that thick deposits contain- 
ing coarse sediments must have been laid down near the margin of a land 
that suffered prolonged erosion. The great thickness of deposits may 
suggest that sedimentation began in an excessively deep basin. How- 
ever, there is unquestionable evidence that nearly all sediments involved 
in mountain folds were laid down in shallow water or at only moderate 
depths. Accumulation of such deposits to a total thickness of several 
miles indicates that slow subsidence of the sea floor was continuous or 
recurrent while deposition was in progress. Moreover, as enormous 
volumes of coarse sediments were delivered into the subsiding basin 
repeatedly, the wasting land must have risen continuously or recurrently 
adjacent to the area of sedimentation. In any case, the preliminary 
structure that determines the location of a future range appears to be a 
sinking trough into which the waste from near-by land accumulates to 
unusual depth. An elongated subsiding tract of this nature is known 
as a geosyndine. Modern examples may be the Great Valley of Cali- 
fornia and the enormous Indo-Gangetic flood plain of India, 

In the Appalachians various features of the strata indicate that 
conditions within the old geosynclinal trough fluctuated repeatedly. 
Sandstones and shales with abundant ripple marks and mud cracks are 
interbedded with thick limestones that contain marine fossils. Such 



relations imply a shifting coast and considerable variation in depth of 
water. In fact at some periods the sea gave place to great delta plains 
or to enormous swamps in which materials for coal beds accumulated. 
These changes depended on the relative rates of subsidence and sedi- 
mentation. If sinking of the trough halted for a considerable time, ac- 



Fig. 271. Map showing the situation of the Appalachian geosyncline and of the old 
land Appalachia. Ad, mass of the Adirondacks. 

cumulating sediments made the sea shallow or even displaced the sea 
water entirely over wide areas. With renewal of subsidence the water 
came back. If the adjacent land was elevated rapidly for a time, erosion 
may have been stimulated sufficiently to keep the seaway full even 
though subsidence of the trough was continuous. 

From study of the sedimentary sections in the Appalachians it is clear 


that the coarser sediments are on the east, and to the west these give way 
to marine shales and limestones. Therefore the land from which sedi- 
ments were eroded lay to the east of the geosyncline. A narrow belt of 
ancient rocks near the present coast presumably represents the western 
edge of the former land; but it must have extended far eastward, over the 
area of the present continental shelf, in order to have volume adequate 
to explain the vast quantities of sediments it supplied. The name 
"Appalachia" is applied to this ancient land of unknown extent (Fig. 
271). The sea that lay west of it, covering much of the present Missis- 
sippi Valley region, was shallow and fluctuating. 

Similar histories have preceded the making of other great ranges. 
The Rocky Mountains grew up from a great geosyncline that stretched 
from the Gulf of Mexico to the Arctic, with highlands to the west. The 
old land that furnished sediments for the Alpine geosyncline lay to the 
north of the Alps. Strata folded in the Caucasus were derived from 
lands to the south while seas stretched northward over Russia. Thus it 
may be accepted as a general principle that on one side of the mountain 
zone lies an area of much older rocks, the source of the folded sedimen- 
tary deposits. The time occupied in the accumulation of sediments in 
the geosyncline extends over long geologic periods. 

Period of Crustal Movements. The period of relatively quiet prep- 
aration, of long-continued erosion and sedimentation and slow move- 
ments of land surface and sea bottom, gives way to a period of greater 
activity in which the Earth's outer shell yields to powerful lateral pres- 
sure. By this pressure the accumulated load of sediments is thrown 
into folds, crushed and mashed together into the disordered arrangement 
characteristic of the great chains. The process and its results thus 
simply stated are in reality very complicated, with different phases and 
with divergent features in different regions, some of the more important 
of which demand separate consideration. We shall take up first the 
operating forces and then the results produced. 

It is clearly evident, from the structural features found in complex 
mountains, that crushing of the old geosyncline and its burden of sedi- 
ments was performed by forces acting in a lateral direction, tangential 
to the Earth's surface. Thus in zones of most intensive folding, the 
folds not only become closed so that their limbs are in contact, but they 
are even more severely compressed, with mashing of the beds and the 
production of very complicated structures. This is illustrated by 
sections in the Alps and in the Appalachians (Fig. 272). Considering 
the scale of these sections it is impossible to imagine the formation of 
such structures except by transverse compression of great magnitude. 

The varied phenomena of folding shown in the mountains may be 


imitated by lateral compression applied to a sequence of artificial strata 
composed of some plastic substance, such as wax or clay, placed upon one 
another. If some of the layers are made somewhat harder and stronger 
than others, and the series is laid in a firm trough or box, one end of 
which may be forced inward by turning a screw, the stronger layers tend 
to form large and simple folds, whereas the weaker members are dis- 
torted in a more complicated way. With continued pressure the folds 
may be overturned and broken. The displacements and dislocations, 
the folding and faulting of the strata, produced in miniature by this 

Fig. 272. Section- across the Santis Alps, N. E. Switzerland (after Heim, somewhat 
modified), a, shales, breccias; 6, massive limestone ; c, shales, thin limestones. 

method, are similar to those observed on a great scale in the mountain 
ranges (Fig. 273). 

Cleavage. It has been shown that cleavage in metamorphic rocks, 
such as slates, is produced by great pressure, and that the planes of 
cleavage are developed at right angles to the direction of pressure. 
Many of the rocks of the great ranges in the zone of intensive folding 
have been turned into gneisses, schists, or slates, depending on their 
original composition and other factors. This alteration becomes more 
evident as the inner portions of the compressed masses are exposed by 
erosion. Observation shows that the planes of cleavage usually stand at 
high angles and not uncommonly are vertical, whereas the strike of the 
cleavage planes is generally parallel to the axis of the range. The 
direction of the compressive force, thus indicated by the cleavage, is the 
same as that shown by the folding. 

Faulting. It is obvious that such extreme folding of rocks could not 
take place without rupturing, breaking, and displacement of the strata. 
We find, accordingly, that the phenomenon of faulting is very common 
in mountain ranges. As we pass from consideration of the simpler 
fold ranges to those of more complex types the faulting becomes more 
pronounced until finally it culminates in thrust faults of enormous mag- 
nitude. The small angle of incidence of the thrust planes to the hori- 
zontal and their trends parallel to the axes of the ranges are indicative of 
the lateral force, or approximately horizontal compression, that has 
produced them. 


In summation, then, we may accept it as a well-grounded fact that the 
structure of the complex ranges has been made by the lateral shoving, 
or squeezing together, of the stratified beds laid down in geosynclines. 

Amount of Compression. The magnitude of the forces involved and 
' of the masses operated upon is indicated by the amount of compression 

Fig. 273. Layers of wax and plaster folded by lateral pressure, imitating structures 
found in mountain ranges. The thrust is from the right, and in successive layers from a 
to e the amount of shortening can be seen. (Willis, U. S. Geol. Surv.) 

which investigation shows has actually occurred in some of the great 
folded belts. In the Appalachians, estimates of 40 to 50 miles, and in 
some sections even more, are given for the distances the original width of 
the strata in the geosyncline has been decreased by the mashing together 
of the mass. In other words, if the folded strata in Pennsylvania, which 
resemble a crumpled blanket, could be smoothed out toward the south- 
east, their extent would be increased sufficiently to cover the state of 
New Jersey. The Rocky Mountain structures represent a comparable 


amount of shortening. Thus the original breadth of the geosynclines 
has been diminished by tens of miles, and in the zones of intensive folding 
and mashing the reduction has been one half or more. 

Influence of Resistant Elements. The old upland along whose margin 
the sediments have been deposited forms a resistant element in the 
architecture of the outer shell. It tends to rise as the geosyncline sinks, 
and as it becomes eroded the stronger massive rocks, igneous and meta- 
morphic, of which its lower levels are composed, are exposed at the sur- 
face. Thus the old land becomes steadily more massive and resistant, 
a more unyielding block or element in the shell. By contrast the sinking 
zone of accumulating sediment is one of weakness; the sinking is pri- 
marily the cause for the accumulation of the sediments, but probably 
the growing load of deposits causes additional subsidence. Finally, when 
the shell yields to lateral compression, the weak sediments are crushed 
against the strong adjoining mass and are crumpled. Rocks in the 
resistant element are mashed somewhat, but there are no stratified 
formations in it to yield by folding. Irregularities in the margin of the 
old land appear to have an influence on the trend of the mountain folds, 
causing prominent bends or arcs. Therefore the general plan of fold 
ranges seems to be determined in an important degree by the situation of 
the old lands, which act as resisting buttresses at the time the folds are 
formed. Thus the Appalachians from southern New York to Alabama 
imitate roughly, in their sinuous trend, the former western coast of the 
old land Appalachia. It is suggested also that the curving trend of the 
Alps has been determined in large measure by old land masses, parts of 
which are now visible in central France, the Vosges, the Black Forest, 
and Bohemia. 

Fig. 274. Diagrammatic section across part of the Jura Range, showing simple struc- 
ture and symmetrical folding. 

Variations of Folded Structure. It is to be expected that the results 
of folding should differ considerably in character between separate 
ranges, or between distant parts of the same range. Differences in 
thickness of sediments, in proportions between strong and weak forma- 
tions, in the form of old rigid masses that transmitted the thrust, and in 
severity of the lateral forces, are reflected in the individuality of moun- 
tain folds. The Jura Mountains, a small member of the Alpine system, 
furnish classic examples of symmetrical, upright folds (Fig. 274). These 
folds were produced far out in front of the Alps proper, in a relatively 


thin sheet of strata, as an incidental effect of the forces that deformed the 
greater Alpine zone. The Appalachians present a wider variety of 
fold structures. In the slate and anthracite regions of eastern Penn- 
sylvania the folds are closely compressed, many of them to the isoclinal 
stage, and the axial planes are strongly overturned toward the northwest. 
Farther west in the state the folds tend to be open and upright; and the 
deformation dies out gradually westward. Going to the south, through 
Virginia and Tennessee and into Alabama, we find that many of the 
folds were ruptured by the severe compression and developed into thrust 
faults (Fig. 275). This kind of complexity is especially pronounced in 
the Alps, which merit special description. 

Thrust Faults and Recumbent Folds of the A Ips. Alpine structure is 
characterized by great folds that have been pushed over to a horizontal 
attitude, and by flat thrusts that are related to these overturned folds. 
These features are developed on an unprecedented scale, with the result 

Fig. 275. Section 12 miles long illustrating Appalachian structure near Greeneville, 
Term. (Slightly modified from Keith and Willis.) 

that the Alps consist of a series of great rock sheets, driven one over 
another and overlapping like the shingles on a roof. The Germans call 
the individual sheets decken; the French refer to them as nappes. 

Because of their location, the Alps have received more intensive study 
than any other mountains. Accordingly, in spite of astonishing com- 
plexity, their structure and history are well known. Like the Appa- 
lachians, they resulted from deformation of thick marine deposits; but a 
large part of the Alpine sediments bears evidence of deposition in deep 
water, far from any shore. Land lay to the north, in the present position 
of central Europe, where mountains of nearly the same date as the Ap- 
palachians were being eroded. Orogenic movement began in Mesozoic 
time, with pressure from the direction of Africa. The soft sediments on 
the sea floor were bowed up slowly, until islands, and chains of islands, 
appeared above sea level. During early Tertiary time the compression 
accelerated powerfully, and an enormous rock sheet was driven north- 
ward over Europe. Beneath this sheet the plastic sediments suffered 
extreme distortion. With recurrent thrusting during the Tertiary other 
sheets were driven forward, and all were severely folded (Fig. 276). 
Erosion cut valleys and "windows" through the sheets, exposing the 
entire series; and in parts of the Alpine area nearly the whole of one or 
more sheets has been swept away, leaving remnants of old rocks to form 



isolated peaks standing on younger rocks that were overridden and 
covered during the thrusting movement. Isolated peaks that have this 


^l^, - 

Fig. 276. Development of Alpine structure. A, block diagram representing part 
of the Alps in an early stage. B to M, cross sections to show successive stages from be- 
ginning of folding (B) to the final intricate structure of thrusts and folds (Jkf). Figures 
show horizontal shifting of corresponding masses. Northwest on the left. (Emile 

anomalous relation are called "mountains without roots." The Mat- 
terhorn and the Mythen are famous examples. Some of these masses 
are 50 or even 100 miles north of their original positions. Heim, the 


great Swiss master of Alpine structure, tells us that the Alpine zone as a 
whole was made narrower by considerably over 100 miles due to the 
thrusting and folding. Locally, as in the Simplon tunnel section, the 
original width was reduced as much as 90 per cent. 

Mountain Elevation. There is a natural tendency to assume that 
mountains rise to greater heights continuously with folding and thrust- 
ing, as a logical result of crowding excess material into a narrow zone. 
For a long time, indeed, this conclusion was taken for granted, and 
attempts were made to compute the original height of eroded ranges by 
determining the amount of shortening due to folding. As the steps in 
mountain history become clearer, however, it is found that much of the 
actual elevation occurred at a distinctly later time than the folding and 
thrusting. After the Rocky Mountain deformation in early Tertiary 
time, the folded and faulted area was eroded to a nearly even surface at 
a low altitude; and the present great heights in the Rockies are due to 
vertical uplift in late Tertiary time. Similarly, after much of the thrust- 
ing and folding was complete the Alps had only moderate height, and the 
sea washed the flanks of the range both on the north and on the south. 
In very recent geologic time a movement of vertical uplift carried the 
Alpine summits to great height. The Andes and the Himalayas have 
had a similar history. 

If the principle of isostasy is kept in mind, it does not seem strange 
that horizontal pressure acting alone fails to force the folded zones to 
great height. An overload results from horizontal transfer and piling 
up of the rocks, and the mountain area continues to subside under the 
increasing weight, in order to maintain approximate balance with adja- 
cent parts of the crust. Therefore even the enormous amount of 
material heaped together in the Alpine zone did not result at once in 
high mountains. Presumably there was slow flowage outward from the 
deformed area, in a deep plastic zone, to prevent an extreme overload. 
The cause of later vertical uplift is a matter for speculation. It is sug- 
gested that the cold rocks of the upper crust, carried to a deeper and hot- 
ter zone by the continued heaping up and sinking, slowly changed their 
state due to heating and expanded. Such increase in volume would not 
change the total weight of the mountain mass, and therefore the surface 
would rise without disturbing isostatic balance. In a qualitative way 
this explanation is satisfactory; but the problem involves many unknown 
factors and therefore cannot be solved quantitatively. 

Rdle of Igneous Agencies. Although the making of mountain struc- 
tures by compression appears to be independent of direct igneous action, 
and some ranges contain little or no visible igneous rock, nevertheless an 
upwelling of magma to produce both intrusive and extrusive bodies has 



been a common incident in mountain making. The effect of this action 
is to modify the structure due to folding and faulting, and by the addition 
of large massive bodies to increase the rigidity and strength of the 
mountain zone. Probably the most effective way in which this happens 
is by the intrusion of great batholiths, usually composed of granite, into 

\\V II Y y y v v v y v v v \V//y 

Fig. 277. Diagram to represent the granitic core of a mountain range after prolonged 
erosion. The granite batholith cuts across folded and metamorphosed sediments. 

the inner, lower portion of the range, A granite intrusion of this nature 
may become exposed later by deep erosion, and is then spoken of as the 
"granite core" of the range (Fig. 277). Intrusion of the heated magma, 
combined with the folding and mashing of the strata, causes profound 
metamorphic effects over wide areas. As great intrusive bodies of this 
kind generally cut across the folds and thrusts, it appears that the in- 
trusion occurred late in the period of orogeny and had no direct con- 

Fig. 278. Section illustrating intrusions of igneous rock (black) in a folded and dis- 
located mountain region. Gr, edge of a granite batholith. 

nection with the compressive forces. Possibly some of the batholiths 
were made by melting of rocks forced into the depths during the folding 
and consequent sinking. Invading granitic masses of this character 
are conspicuous features of the Coast Range in western Canada, of the 
Sierra Nevada, the Green and White mountains in New England, the 
Caucasus, and the mountains of Scotland and Norway. 

Intrusions of molten magmas may not only make great batholiths, but 
pressing upward along belts of weakness caused by folding and faulting, 
they may form intrusive sheets, laccoliths, dikes, and other bodies (Fig. 
278) ; or, attaining the surface, they may be extruded as lava flows or 


give rise to violent volcanic action. During the formation of the Rocky 
Mountains, Wyoming and Montana were the scene of great volcanic 
activity, the dying phases of which are still evident in Yellowstone Park. 
As a result of the folding and faulting of strata, and the intrusion and 
extrusion of magmas, there were produced ranges with geologic struc- 
tures of wonderful complexity, now revealed to us by deep dissection. 

Subsequent Stages. Although mountains may be classified accord- 
ing to the crustal movements involved in their formation, it is not to 
these processes alone that the mountains in their present form are dlie. 
Hand in hand with uplift of the masses goes the work of erosion, that 
mighty chisel of Nature, which shapes and carves them into the forms 
familiar to us. In a sense, therefore, all mountains are residual, as 
erosion starts with the beginning of uplift, proceeds during the whole 
period of orogenic activity, and continues so long as the highlands exist. 
During the subsequent stages the history of a range consists chiefly of 
progressive changes due to erosion, by which the mountains reach 
maturity and old age. 

Crustal movements in the mountain zone do not cease entirely, how- 
ever, at the time the ranges reach their maximum height. As erosion 
proceeds its work is partly offset by recurrent up warping of the region. 
Old mountains rejuvenated by a strong uplift are sometimes called 
'posthumous mountains. The old Appalachians have been bowed up 
repeatedly, even in recent geologic time. After their birth during the 
Permian they suffered erosion through long geologic periods; and al- 
though uplifts probably occurred, the entire folded tract finally was 
reduced to a peneplain. During the Tertiary the region was warped up 
strongly and dissected; and therefore the present mountain ridges are 
strictly residual mountains. The latest pulses of uplift are recorded-by 
high terraces along the streams. 

This tendency of mountain, ranges to renew their growth may be con- 
nected with isostasy. After a large volume of eroded debris is trans- 
ferred from the mountains to the plains or the ocean basins, the moun- 
tain mass is forced to rise to restore equilibrium. It does not rise to its 
original height, however, probably because the plastic rock forced in at 
the base of the rising mass has a higher density than the rock eroded 
from the surface. In time, therefore, by removal of an enormous volume 
of material, the mountains may be reduced permanently to a low level. 


The geologic period in which a geosyncline has been crushed; and the 
strata compressed into mountain f olds, is fixed by determining the age of 
the latest strata involved in the folding, and of the oldest undeformed 


beds that lie upon the disturbed rocks about the mountain flanks. 
Obviously the folding is younger than any of the folded or disturbed beds 
and older than any that are undisturbed (Fig. 279). The closeness of 
dating by this method evidently depends upon the length of the interval 
between the dates of deposition of the two sets of strata. Thus if the 
youngest folded rocks are of late Triassic age, and the oldest undeformed 
sediments were deposited early in the Tertiary, the folding may have 
occurred either in the Jurassic or in the Cretaceous period. The Ap- 
palachian folding is dated rather closely, as Permian rocks are affected 
and Triassic rocks are not. 

As many mountains have been formed not in one, but in several periods 
of compression, the method just explained can be used to date only the 
latest movement. However the earlier disturbances may be recorded 

Fig. 279. Illustrating determination of the date at which deformation in a mountain 
zone occurs. A are the youngest surviving strata that were involved in the folding; 
the folding is younger than these. B are not concerned in the process, and hence the 
folding is older than they are. 

by unconformities. Thus if a surface of unconformity cuts across 
strongly folded rocks, and the formations above the unconformity also 
are deformed, there must have been two distinct epochs of orogeny. 
The several pulses of movement that affected the Alpine zone are dif- 
ferentiated by evidence of this kind. 

Mountain making by any process is dated according to the same gen- 
eral principles. The latest episode of uplift in fault mountains cannot 
be older than the youngest rocks affected by the faulting. 


When it was believed that the Earth consisted of a relatively thin 
crust resting on a highly-heated liquid interior, it seemed easy to explain 
the origin of mountains by assuming that there was a regular contraction 
of the Earth's mass from loss of heat, and, as this contraction was 
greater in the heated interior than in the cold outer crust, the latter was 
folded up as it gradually sank upon the shrinking core, very much as the 
skin wrinkles upon a drying and contracting apple. This view can no 
longer be held, because the Earth is known to behave as a solid, rigid 
body; the interior cannot be wholly, or even chiefly fluid, in the ordinary 
meaning of the word. 


Nevertheless a view commonly held to account for crustal deformation 
assumes contraction by loss of heat. This is a survival of the idea men- 
tioned above, but changed to accord more nearly with later knowledge. 
It assumes that the Earth is solid and rigid but very hot within, and that 
progressive loss of heat causes slow shrinkage below a comparatively 
shallow depth. An English scientist has shown that this mechanism 
would account for a large amount of lateral thrusting in the outer shell 
throughout geologic history, provided we make certain reasonable as- 
sumptions as to the temperatures in the interior. Another form of the 
contraction hypothesis discards the idea of cooling, but assumes that the 
enormous pressures deep in the Earth cause matter to increase in density 
and decrease in volume. The net effect would be the same as if cooling 
occurred, as the outer shell, unchanged in volume, would collapse on the 
shrinking core, with consequent folding and thrusting. In either form 
the hypothesis encounters grave mechanical difficulties; among them, the 
necessity for explaining localization of deformation, during anyone period, 
in only two or three widely separated geosynclines. It seems more prob- 
able that uniform contraction of the Earth would cause moderate failure 
of the crust in many zones, instead of violent deformation in a few. 

It has been suggested that vertical adjustments in the outer crust to 
maintain isostatic equilibrium may be a sufficient cause of deformation. 
This suggestion does not find support in the evidences of enormous lateral 
pressure and displacement. In some mountain belts individual rock 
sheets have been thrust horizontally for distances of 25 miles or more; and 
every large folded geosyncline represents lateral movement through tens 
of miles. No force acting vertically in the crust could have a horizontal 
component so large. Moreover the fact of concentrated deformation in 
a few long, narrow belts is as difficult to explain by isostasy as by the 
contraction hypothesis. 

Within the last few years some geologists have suggested that whole 
continents may shift horizontally through long distances. It is claimed, 
for example, that Africa moved northward against the old Mediterranean 
geosyncline and crushed it to form the Alps and neighboring mountains; 
and that the great folded chains of Asia were caused by southward 
shifting of that continent. It is urged that no other explanation will 
suffice in view of the stupendous shortening recorded by mountain folds 
and thrusts. But even if we should admit the moving of continents, the 
fundamental problem of orogeny would remain unsolved so long as the 
ultimate forces and conditions to cause such movement are wholly 

It must be admitted, therefore, that the cause of compressive deforma- 
tion in the Earth's crust is one of the great mysteries of science and can 


be discussed only in a speculative way. The lack of definite knowledge 
on the subject is emphasized by the great diversity and contradictory 
character of attempted explanations. It is a fascinating problem, but 
lengthy discussion of its various aspects has no place in this volume. 
The facts and relationships of mountain structure present a large field of 
study in themselves, aside from the problem of ultimate forces. 

Fault mountains that are due chiefly to thrust faulting present prob- 
lems similar to those encountered in mountains with folded structure, as 
the fundamental cause is enormous lateral compression. Some steep 
faults that bound mountain blocks appear to be related to irregular 
shifting of magmas at depth, during the formation of intrusive or ex- 
trusive igneous bodies; and others are explained by vertical movements 
to restore or maintain isostatic balance with shift of load. Still others 
are formed in or near belts of severe folding, by irregular local twisting of 
the crust during the period of compression; and as fault mountains made 
in this way are merely incidental effects of the larger deformation, they 
become a part of the greater problem. Many dome mountains certainly 
owe their formation to igneous intrusion, and therefore the problem of 
their ultimate cause is closely linked with the problems of igneous 
activity. Some of the larger dome mountains appear to be closely re- 
lated in origin to great folded units, and therefore involve similar un- 
certainties as to origin. 


Mountains may be classified generally according to types of structure, 
which reflect different processes and forces that acted on the mountain 
zone. The visible structure may be due to simple dislocation and tilting 
of blocks, to simple doming of rocks, to folding with or without faulting, 
to large thrust movements, or to various combinations of these diverse 
processes. Connected with any type of movement there are commonly 
injections and extrusions of igneous material, which complicate the final 
structure of the mountain mass. Nearly all mountain-building move- 
ments take place in a series of pulses or phases distributed over a very 
long time. This is especially true of the chains that have complex 
structure. The cause of the great lateral pressure to which the belts of 
folding and thrusting bear eloquent testimony is an unsolved problem. 
We only know that long belts of weakness in the crust indicate their 
presence first by continued subsidence as geosynclines, and finally by 
yielding to forces that make the structures characteristic of the great 
mountains. Actual uplift to mountain heights, however, commonly 
follows the period of folding and appears to be largely independent of it. 


Erosion is active in mountains throughout their history, and leads to 
progressive evolution of their varied forms. This aspect of mountain 
history is discussed systematically in the following chapter. 


1. Mountains: Their Origin, Growth, and Decay; by James Geikie. 285 pages. 
D. Van Nostrand Co., New York, 1914. 

Not up-to-date on all points, but has good descriptions, well written. 

2. The Structure of the Alps; by L<k>n Collet. 272 pages. Edward Arnold & Co., 
London, 1927. 

An excellent guide book for the study of Alpine geology in the field. 




Extended Streams. It has already been shown (Fig. 163) that 
when the sea encroaches upon a subsiding land mass, stratified de- 
posits of different kinds are laid down with gentle seaward inclination. 
During the maximum submergence the shore deposits rest against 
the partly submerged oldland. When the sea has retreated from the 
area, the result is an uplifted coastal plain built of gently-dipping sedi- 
mentary rocks. The plain is not unbroken, however, because the 
streams of the oldland must have extended themselves mouthward as 
the sea began to retreat, in order to keep pace with the retreating shore- 
line, cutting at the same time an ever lengthening trench in the newly 
exposed coastal-plain rocks. Such streams are known as extended 
streams. They are of course consequent upon the uplifted surface 
as well (page 73); hence they are called extended consequents (Fig. 

Subsequent Streams ; Cuestas. As rain falls on the exposed coastal 
plain, tributaries develop, working headward from the trenches of the 
extended consequents. They develop most readily along zones of weak- 
ness in the underlying rocks. Such a zone is found in the contact of the 
recent shore deposits upon the much older, harder rocks of the oldland. 
Along this former land margin, then, the tributaries rapidly develop, 
working headward in a direction at right angles to that of the main 
streams. Such tributaries are called subsequents; all streams that work 
headward in this way along weak rock belts are so called. The parallel- 
ism of the weak belts in this case causes the subsequents to enter the 
main streams at right angles. The rectilinear drainage pattern thus 
developed is called a trellis pattern. (Compare the dendritic pattern 
developed in homogeneous rocks, page 76.) These subsequent streams 
cut the seaward, soft-rock sides of their valleys much more rapidly than 
the oldland sides where the rock is hard. Hence their channels con- 
stantly tend to shift seaward as by erosion they etch away the edge of 
the uppermost coastal-plain stratum. The product of this etching is a 
strongly asymmetrical low ridge, with a steep escarpment facing land- 
ward and a very gentle slope facing seaward formed by the surface of the 




gently dipping coastal-plain deposits (Fig. 281). Such a ridge is known 
as a cuesta (the Spanish word for this type of land form). 

Belted Coastal Plain. Wherever the extended consequents discover 
a resistant layer overlying a weaker layer in the sedimentary series, they 


Pigs. 280-282. Development of relief on a newly uplifted coastal plain. 

Pig. 280. Newly uplifted coastal plain composed of gently dipping beds of unequal 
resistance and showing progressive overlap, and drained by extended consequent streams. 

Pig. 281. Etching out of a cuesta from the topmost resistant bed by the development 
of subsequent streams. 

Pig. 282. Evolution into a belted coastal plain by the development of additional 
cuestas as lower strata are exposed. 

send out subsequents which cut into the underlying soft rock, undermine 
the hard cap, and etch it slowly down its dip toward the sea. In this 
way a whole succession of cuestas may be developed, separated by the 
troughs of subsequent streams. Coastal plains bearing two or more 
cuestas are called belted coastal plains (Fig. 282). 

The escarpment of a cuesta increases in height until the main conse- 
quents have become graded. At this stage downcutting by the tribu- 


taries has virtually ceased, and since they can thus cut only laterally, the 
cuesta escarpments are gradually lowered by being etched seaward, until 
they disappear. The result is a peneplain. 

The coastal plain of New Jersey is belted, one of the belts being formed 
by the Navesink Highlands. In southern England, the North Downs 
and South Downs form two well developed series of cuestas facing each 
other across the Weald, a broad gentle anticline. Much of the country 
between Paris and the eastern frontier of France consists of a series of 
cuestas with east-facing escarpments. This arrangement was of the 
highest importance to the Allied armies on the western front during the 
Great War, since it presented a series of difficult natural obstacles to the 
German advance. 


Under humid climatic conditions, the denudation of an area of folded 
strata follows exactly the same laws as in the case of horizontal or 
homogeneous rocks outlined in Chapter IV. Local complications, 
however, are introduced because of the unequal resistance to erosion of 
alternating hard and soft beds. A typical case is outlined below which 
in many respects is similar to the early erosional history of the folded 
Appalachian ranges. Many variations of course are possible, but the 
following account is representative of the cycle of erosion and deposition 
under these conditions. In connection with this account, reference 
should constantly be made to Figs. 283-288. 

Initial Stage. Let the scene be set with a land mass in old age (es- 
sentially a peneplain) developed on a series of horizontal sedimentary 
rocks, alternately resistant and weak, of which two upper resistant beds 
with included weak beds are to be later affected by erosion (Fig. 283). 
The area is drained eastward by one main stream which reaches the sea 
over a well-developed flood plain and delta. The tributaries, because 
the rock is horizontal, form a typical dendritic pattern. 

Stage of Early Youth; Antecedent Streams and Structural Valleys. 
As folding begins, this section of the crust is thrown into three gentle anti- 
clines (Fig. 284), of which the first plunges northward. The increased 
gradient attendant upon the folding brings the sluggish main stream to 
life, greatly increases its erosive power, and allows it to trench downward 
as rapidly as the anticlinal arches are uplifted. A stream which is strong 
enough thus to maintain its former course in spite of gradual uplifts 
across its path is "said to be antecedent. Columbia River is anteced- 
ent to the great uplifted arch of the Cascade Range in Oregon and 
Washington. The tributaries, however, smaller and weaker than their 


Figs. 283-288. Ideal cycle of stream erosion under a humid climate and in folded 
strata. Each block is oriented so that its right end faces east and its long upper edge 
faces north. 

Fig. 283. Initial stage, showing a land mass in old age. 

Fig. 284. Stage of early youth, showing the beginning of folding and the develop- 
ment of an. antecedent main with consequent tributaries in a trellis pattern. 

Fig. 285. Stage of later youth, showing valley fill, extensive breaching of the anticlines, 
and the development of subsequent valleys bordered by hogbacks. 

Fig. 286. Stage of maturity, showing dissection of the earlier valley fill, decrease in 
relief, and the development of rounded ridges within the anticlines. 

Fig. 287. Stage of old age, showing the weak rocks reduced nearly to baselevel, leaving 
only monadnocks of resistant rock for the streams to cut. Essentially a peneplain. 

Fig. 288. Stage of maturity in a second cycle instituted by regional uplift and re- 
juvenation. (See page 414.) 



nain, fare less well; they are blocked off by the folds and are extin- 
guished. But new drainage lines are provided by the synclinal 
troughs. Concentration of drainage takes place in each of these, and 
pairs of tributaries are thus developed at right angles to the main stream, 
as consequents upon the folded surface. In this way the dendritic 
stream pattern is extinguished and a right-angled, or trellis pattern, takes 
its place. The synclinal tributary valleys are termed structural valleys 
because their position is determined entirely by rock structure. At the 

Fig. 289. Lehigh Water Gap as seen from Slatington, Pa. (Barrell.) 

same time, rapid run-off down the slopes of the anticlinal limbs forms 
consequent tributaries of the second order, which rapidly excavate deep 
gorges. Directly these tributaries cut through the surface bed of re- 
sistant rock, and penetrate the weak bed below, they begin to cut the 
latter, undermining the former, and the valleys are quickly widened. 
This process is characteristic of the early stages of the cycle, in that the 
weak rocks are attacked as widely as possible, while the hard rocks, 
escaping direct attack, fall by undermining. The main stream crosses 
the anticlinal arches through water gaps of its own cutting (Fig. 289), and 
the greatly increased volume of debris in transport is deposited in part 
upon the uplifted coastal plain. That the stream system is confronted 
by a long period of cutting is indicated by the volume of rock above the 
new baselevel (indicated by a dashed line, Fig. 284). 

Stage of Later Youth; Hogbacks and Canoe-shaped Valleys. In 
this case the folding and general uplift is depicted as having now reached 
its climax (Fig. 285). The crests have thereby attained their maximum 



elevation above baselevel (dashed line, Fig. 285). But because much 
erosion has taken place since the uplift began, it is clear that the actual 
mountains can never be as high as might be inferred at a later time by 
projecting upward the stumps of the anticlinal limbs. Steepening of the 
anticlinal slopes (compare Fig. 285 with Fig. 284) has resulted in such 
greatly increased erosion by the tributaries of the second order that the 
streams in the synclinal valleys, whose gradients have not been steepened 
much since folding began, are not able to carry away the waste contrib- 
uted to them. The synclinal valleys are therefore partly silted up with 

Fig. 290. A hogback near Gallup, New Mexico. Buggy near left base gives scale. 

(U. S. Geol. Surv.) 

waste in the form of coalescent alluvial fans. The ridges formed by the 
outcropping hard bed resemble steep cuestas. Such ridges of steeply- 
dipping rocks are termed hogbacks. They are common in the Colorado 
Rockies, the Bighorn Mountains, the Black Hills, and many other anti- 
clinal mountain masses (Fig. 290). Whereas the anticline in the center 
(Fig. 285) has been breached along its entire crest, the anticline to the east 
is only partially breached, and that where its axis is highest. Since this 
is a plunging fold, its axis descends to the north, and hence the tributaries 
of the second order have lower gradients. This being the case, their 
cutting power is far less than that of their steeper neighbors, and the 
resistant bed which caps the anticline is therefore much less readily cut 


through. This brings about the excavation of a canoe-shaped valley r , 
the prow of the canoe pointing down the plunge of the anticlinal axis 
whereas its sides are formed by the anticlinal limbs. The canoe gradu- 
ally widens as the retaining escarpments retreat. The whole gaunt 
structural skeleton is now dissected out in greatest relief, and the stream 
system has much cutting to accomplish before it can remove all of the 
rock above baselevel and carry it into the sea. 

Stage of Maturity. The weak strata continue to receive the most 
vigorous direct attack by the streams, and the resistant strata fall block 
by block as they are sapped and undercut. Both the canoe-shaped 
valleys and the others cut in the anticlines have been widened thereby 
(Fig. 286). Erosion of the soft strata in the middle anticline has now 
completely laid bare the rounded back of the lower resistant bed, which 
has already been breached to form an inner canoe-shaped valley. In- 
spection of Figs. 285 and 286 shows the appearance and development 
of other such rounded ridges at the hearts of the anticlines. 

The whole land surface by this time has been reduced appreciably 
toward baselevel (dashed line, Fig. 286; compare Fig. 285). In fact, the 
heights have been lowered sufficiently to decrease greatly the supply of 
waste to the tributaries. The latter, freed from a part of their load 
of sediment in transport, are able again to devote a part of their energy 
to downcutting, and they begin to dissect their previous deposits in the 
synclinal valleys and on the coastal plain, cutting them into pronounced 
terraces. The main stream has now come so close to baselevel that it can 
cut down but little further. Hence it is easily diverted, and it begins to 
meander widely, broadening its valley by lateral planation. In this it is 
slowly followed by the tributaries in the synclines. 

Stage of Old Age ; Beveling of Resistant and Weak Rocks Alike. 
The increasing predominance of lateral planation during late maturity 
and old age has resulted in the virtual cutting away of the remaining 
highlands. The reason is simply that all of the soft rock has already 
been worn down essentially to baselevel, leaving nothing but hard rock 
for the streams to cut (Fig. 287; note that the dashed line has disap- 
peared because it practically coincides with the old age surface). Be- 
cause of the great resistance of the only erodable rock and the low gradi- 
ents, the process of cutting is almost infinitely slow. Figure 287 depicts 
a peneplain carrying elongate monadnocks formed by outcropping edges 
of the resistant beds, together with one rounded ridge in the plung- 
ing anticline. The structure is essentially beveled, hard and soft alike. 
The cycle is virtually ended; from now on the streams can work only to 
destroy the monadnocks, with scarcely appreciable results. 



1. Downward Movement; Drowning. Downward warping of the 
crust beneath any part of a stream's course causes a local decrease in 
gradient which may be followed by extensive deposition. The effects of 
downward movements are most plainly visible near the mouths of 
streams, which, if their channels are depressed below baselevel, are said 

Fig. 291. A river drowned in its lower reaches, forming a branching estuary. 
is slowly being destroyed by encroaching deltas. 

to be drowned. Most of the larger rivers of the Atlantic Coast of the 
United States are drowned at their mouths. The lower Hudson oc- 
cupies a deep narrow drowned valley cut in resistant rocks, and Chesa- 
peake Bay represents the drowned lower course of the Susquehanna and 
the Potomac combined. The bay is broad and shallow because cut in 
weak rocks. The mouths of streams in southwestern Britain and north- 
western France likewise are drowned over wide areas. Typical estuaries 
formed by the submergence of a valley are illustrated in Fig. 291. 

2. Uplift ; Rejuvenated Streams and Entrenched Meanders. Up- 
warping of the crust is likely to increase the cutting power of the streams 
affected by increasing their gradients. Streams thus affected are said to 



be rejuvenated. The effects, however, upon the aspect of a valley will be 
far more noticeable in the case of an old or mature stream rejuvenated, 
than in the case of one rejuvenated when still in its youth (Fig. 292). 
This follows from the fact that renewed cutting normally trenches a new 
V-shaped valley in the floor of the old one, and that if the latter is already 



Fig. 292. Cross profiles, of valleys showing that the effect of uplift and rejuvenation 
depends on the stage at which it occurs. 

A Old stream rejuvenated: effect very strongly marked. 

B Mature stream rejuvenated: effect strongly marked. 

C Young stream rejuvenated : effect barely perceptible in the spurs along the valley 

D v el y young stream rejuvenated: effect not registered, (Drawn by E. J. Lees.) 

V-shaped, its appearance in cross profile is scarcely altered. A keen 
eye can sometimes detect the evidence of uplift in a valley still youthful 
(Figs. 292 C, 293). 

If a meandering stream is affected by uplift so rapid that the rate of 
lateral cutting by the stream cannot keep pace with it, the stream will be 

Fig. 293. East end of Boulder Canyon, southern Nevada. Note abrupt changes in 
slope of valley walls, suggesting at least two rejuvenations. (U. S. Geol. Surv.) 

rejuvenated in the exact meandering course it followed immediately 
prior to the uplift. The meander loops thereby come to occupy gorges 
separated by steep-walled projecting spurs (Fig. 294). These loops are 
called entrenched meanders, and are a positive indication that uplift has 


taken place in the region where they occur. Furthermore, the depth of 
the gorges is an approximate gauge of the amount of uplift. 

The meanders cannot remain deeply and narrowly entrenched for 
long. The process of sweep is operative here as in the case of normal 
meanders (page 56). Lateral cutting is greatest against those sides of 

Fig. 294. Entrenched meandering valley of the San Juan River, 30 miles below Bluff, 
Utah; canyon 1200 feet deep. (Vinson.) 

the curves that face up against the general course of the river, since 
upon them the force of the current must come with attendant corrasion. 
These sides of the spurs interlocking between the meanders therefore 

Fig. 295. An entrenched meandering valley; arrows point downstream. C, C, C, 
undercut slopes of valley; S, S, S, slip-off slopes; P, P,P, flood-plain scrolls. (After 

tend to be undercut and to present steep and even cliff-like faces to tne 
course of the river (C, (7, C, Fig, 295) . The down-valley sides of the spurs, 
S, S, S, on the contrary are apt to descend to the river with more or less 
gentle and gradual slopes, commonly covered with sand or gravel de- 
posited by the stream. The reason for the difference is that since the 


river cuts laterally against the faces C, C, and also vertically downward, 
these sides of the spurs are being eaten into and consumed, whereas it 
tends to move away from the sides S, S 7 and with slack current to leave 
deposits on them. Since the stream tends to slide away from them with- 
out eroding, they are called slip-off slopes, as distinguished from the 
undercut slopes C, C. As one looks down the course of such a valley he 
sees only the steep and wooded or cliff-like faces of the undercut spurs, 
which give it a stern and rugged aspect; when he looks up the valley the 
cultivated fields of the gentle slip-off slopes confront him. 

From what has been said above it will be seen that an entrenched wind- 
ing river tends to become more circuitous in its course, and that the 
whole system of meanders moves down the valley. Further, when 
downcutting decreases and the stream becomes graded, ,it begins to 
build narrow strips of land or scrolls, P, P, P (Fig. 295), along the insides 
of the curves S 3 S, 8. Material is gradually and steadily added to these 
scrolls, and they grow in size. As the valley widens by lateral planation 
they coalesce, and a continuous alluvial flat is established. 

It is evident that as the meanders enlarge and change their curves 
they are likely to form short cuts, as shown in Fig. 26. In an entrenched - 
meander it may happen that the neck of land connecting the spur end is 
actually undercut at the narrowest place, leaving the former spur rem- 
nant as an island joined to the mainland by a natural bridge, under 
which the stream runs in the short cut it has made. 

The rivers of northeastern France and Belgium, such as the Meuse 
and the Moselle, furnish typical examples of entrenched meanders. The 
headwaters of the Susquehanna in Pennsylvania, the Kentucky, and 
many others, are equally good examples. 

The completion of the ideal cycle outlined on pages 409-413 and de- 
picted in Figs. 283 to 287 is followed in Fig. 288 by uplift and rejuvenation. 
In other words, the former baselevel (upper dashed line) is abandoned in 
favor of the new baselevel (lower dashed line). This permits the stream 
system to renew the work of excavation which it had completed under the 
old conditions. The weak rocks are again etched away from the resistant 
rocks, which are left standing as prominent ridges. The tops of these 
ridges are beveled flat, and they are worn down so slowly that during all 
of the early part of this new (second) cycle they preserve the level of the 
former peneplain from which the intervening weak rocks have been cut 
away. These ridges are said to be peneplain remnants, and can be formed 
only by uplift and the beginning of a second cycle following the comple- 
tion of an earlier one. In an analogous manner, the highest ridges of the 
folded Appalachian ranges are remnants of a former peneplain. Their 
flat beveled tops betray the fact that they once stood at or close to 



baselevel and that they were later etched into relief following a great 

In Fig. 288 the weak rocks in the two synclines are temporarily pre- 
served as ridges because protected by narrow caps of resistant rock. The 
latter however must disappear in time by being undermined, and the 
synclinal ridges will then be quickly destroyed. The completion of the 
second cycle will see a baseleveled surface (developed in the plane of the 
lower dashed line) much like that of Fig. 287. There will be two chief 
differences, however: (1) Monadnocks will be fewer because of the 
complete removal of the upper resistant layer from the two synclines. 
(2) The remaining monadnocks will be farther apart because the hard 
layers will be intersected farther down their limbs by the plane of the 
new baselevel. Thus the hills, strong bastions in appearance, are in 
reality slowly moving down the dip of the resistant strata of which 
they are composed. 

Rock Terraces Caused by Successive Uplifts. Rock terraces caused 
by differential erosion of horizontal strata are discussed in Chapter IV. 
Terraces are also not infrequently cut in tilted or completely folded 
strata, beveling them regardless of their unequal resistance (Fig. 296). 
Since hard and soft beds alike can be brought down to a common plane 

Fig. 296. Rock terraces caused by successive uplifts. (After Wright. Va. Geol. 
Survey.) Terraces a, 6, and c, are remnants of old graded valley flats, elevated by three 
successive uplifts, the last of which is causing the excavation of the trench d. Width of 
cross section is of the order of magnitude of \ mile. (Drawn by E. J. Lees.) 

only near baselevel, it follows that each pair of rock terraces of this type 
must have been formed by a stream in the latter half of its cycle, and 
that for each pair of similar terraces there must have been a corre- 
sponding uplift of the land followed by renewed downcutting toward 
baselevel and then widening of the valley by lateral planation. Thus in 
Fig. 296, a nearly baseleveled surface a must have been uplifted and 
dissected, followed by valley widening near a new baselevel b. A second 
uplift resulted in the cutting of the plane c which after a third uplift was 
dissected into terraces by downcutting of the trench d. Obviously 
peneplanation took place in no case; if it had, the higher terraces would 
have been destroyed. Each terrace is veneered with a thin, layer of 



alluvium, the remains of an alluvial flat built up by continuous deposition 
on the inside curves of meanders. 

Adjustment of Streams to Structure. A close study of Figs. 283-286 
shows the remarkable fact that although the initial tributaries develop 
as consequents in the synclines, they steadily decrease in importance 
until in old age (Fig. 286) they are almost completely extinguished. The 

50 mi. 

Fig. 297. Outline map showing stream patterns in West Virginia. In the central 
and western regions the rocks are nearly horizontal and the pattern is dendritic. In the 
east the rocks are strongly folded along NE-SW axes and the resulting pattern is treiiised. 
Note how thoroughly the region is drained (mature dissection under a humid climate), 
and compare Fig, 62, Chap. IV. (Drawn by D. Gallagher.) 

further fact appears that as the resistant caps of the anticlines are 
breached, subsequent streams (page 407) are developed in the weak 
rocks below, parallel to the synclinal consequents. As the cycle pro- 
gresses these subsequents gain in importance, so that before the end they 
have become the most important tributaries. This growth of the sub- 
sequents at the expense of the consequents is called adjustment to struc- 
ture and is explained by the simple fact that the former can cut more 
rapidly in thin weak rock beds than the latter are able to cut in the 


resistant rocks through which they flow. From this the universal 
principle may be set up that stream systems are constantly forced to adjust 
their courses so as to flow as much as possible on weak rock and as little as 
possible on resistant rock. Until this has been achieved within the limits 
of existing conditions, the streams are not completely adjusted. Drain- 
age developed in folded strata rarely reaches an adjusted condition before 
late maturity or old age. The more or less complete adjustment of the 
larger streams in the folded Appalachians is illustrated in Fig. 297. From 
the trend of the trellised stream pattern the strike of the folds can be 
readily inferred. 

Superimposed Streams. Certain streams are so completely out of 
adjustment and pursue courses so wholly regardless of structure that 
some special set of conditions must have brought about this inharmonious 
relationship. Suppose the seaward portion of a peneplain developed on 

Pig. 298. Development of a superimposed river. A, course determined by initial 
slope on a layer of sand and gravel ; B, latter removed by erosion and river pursuing its 
course without regard to underlying rock structure. 

folded strata of unequal resistance to be depressed slightly. During the 
resultant submergence, marine deposits are laid down almost horizon- 
tally upon the drowned erosion surface, forming an angular uncon- 
formity (page 332) with progressive overlap (page 234). When now the 
monotonous submarine plain is uplifted, consequent or extended streams 
develop on the gently sloping surface, exhibiting a dendritic drainage 
pattern (Fig. 298, A). In time the streams cut down through the un- 
conformity into the different material and structure below. When the 
overlying sediments have been stripped away, the streams are found to 
be following unadjusted courses inherited from the overlying unconform- 
able strata (Fig. 298, B). Such streams are said to be superimposed. 
Streams may be thus "let down from above" from glacial, volcanic, 
eolian, and alluvial deposits as well as from marine deposits. In any 
case, gradual adjustment will take place as the cycle progresses. 

From the foregoing, it must be true that any stream which cuts across 
the structure of folded rocks, and which therefore cuts water gaps across 
tilted resistant strata, must be either antecedent or superimposed. Neither 
consequents nor subsequents could occupy such positions without first 



becoming antecedent or superimposed. If remnants of overlying un- 
conformable strata still cap the ridges through which the gaps are cut, 
the case for superimposition is fairly clear. Evidence of antecedent 
history is more obsciire. 

Wind Gaps ; Capture by Subsequent Streams. A water gap (Fig. 289) 
abandoned by its through-flowing stream is called a wind gap (Fig. 299). 

Fig. 299. Wind Gap in Kittatinny Mountain at Pen Argyl, 11 miles north of 
Easton, Pa. (Barrell.) 

Most of the wind gaps which have been studied are the result of capture 
by subsequent streams. This process is best illustrated by the case of 
Snickers Gap, a prominent notch in the Blue Ridge of Virginia about 

Fig. 300. A, former drainage across the Blue Ridge in northern Virginia. B, 
present drainage resulting from the beheading of Beaverdam Creek by the Shenandoah. 

fifteen miles south of Harpers Ferry. The floor of this notch hangs 700 
feet above Shenandoah River at the west base of the ridge. A small 
stream called Beaverdam Creek heads near Snickersville at the east 
base, and flows eastward away from the ridge. The notch is a large one, 
and must have been cut by a large stream. Its reconstructed history is 
as follows: During an earlier cycle than the one now in progress, when 
the land around the base of the Blue Ridge stood as high as the notch of 


Snickers Gap, this notch was occupied by a transverse stream, the 
ancestral Beaverdam Creek (Fig. 300, A}. But the much larger trans- 
verse ancestral Potomac, which crossed the ridge 15 miles farther north, 
was able to cut its gap downward into the hard ridge-forming rocks 
much more rapidly. The Shenandoah, tributary to the Potomac through 
an easily eroded limestone area west of the ridge, kept pace with its 
master stream, and with the resulting favorable gradient, worked head- 
ward toward the south, cut through the divide which separated it from 
Beaverdam Creek, and "beheaded 7 ' the latter by diverting its upper 
waters toward the Potomac (Fig. 300, B). The gap was forthwith aban- 
doned, and the beheaded and weakened creek found itself confined to 
the area east of the Blue Ridge. As the Shenandoah continued to cut 
downward, the ridge was etched out in greater relief and the abandoned 
gap was left high and dry. 

Some wind gaps may have been formed by relatively weak streams 
which, affected by upwarps athwart their courses, were able to maintain 
their courses only for a time, having been later forced to abandon them 
and to flow along the strike. 


It has already been pointed out (Chapter XVI) that the work of ero- 
sion goes forward hand in hand with the uplift of great crustal masses, 
etching out forms of mountainous size. The work of erosion, then, is of 
the greatest importance in a full consideration of mountains. It begins 
with the first rising of the masses, proceeds while the orogenic forces are 
at work, and continues long after they have come to rest. As its results 
become especially marked in this last stage, it must be considered the 
chief agent in mountain development during the later phases of mountain 

Earlier Stages of Erosion. So long as the compressive orogenic 
forces are at work, a mountain range grows in so far as its structure is 
concerned. Whether it actually rises in height or not depends on the 
adjustment between (1) vertical uplift which tends to make it rise, and 
(2) the work of erosion which tends to cut it down. Always during the 
formative period this struggle goes on, and the height of the range at 
any time is a function of these two forces. When the orogenic move- 
ments cease, then denudation has full sway and, ultimately, with the 
lapse of time and provided no renewal takes place, the range must be 
cut down, baseleveled, and extinguished by the relentless agents of 
erosion. In this process various stages are to be distinguished. When 
the range is at its maximum elevation the erosive agencies are most 


severe; to the work of running water on steep slopes is added very com- 
monly the effect of frost, snow, and ice. 

It may happen also that at this time the rock material exposed to 
erosion consists of the later beds laid down in the geosyncline, which have 
suffered less metamorphism than the deeper, older ones, and are thus 
less resistant to erosive attack. If igneous extrusions have contributed 
to swell the volume of the range, it will also be the more easily eroded 
tuffs and lavas that are first exposed. Hence, in general, the outer 
material is more easily cut away, and the inner core progressively expo^s 
to erosion more and more resistant rock. Thus, in the early history of a 
range not only is the severity of attack of eroding forces likely to be 
increased by great height, but they may find less resistant material to 
work upon. At first the upraised masses begin to be trenched by the 
valleys of the initial consequent streams. The drainage lines thus ap- 
pear upon original slopes and continue to cut downward and to work 
backward into the range. As they do so they begin to be conditioned 
more and more by the structure and nature of the underlying rocks. 

Fig. 301. Longitudinal profile of a mountain range in early maturity. 

The mountain masses are profoundly graved and acquire rugged peaks 
and towering rock pinnacles, alternating with deeply scored valleys. 
The strongly notched outlines of such ranges present a saw-toothed 
appearance, which has led to their being called by the Spanish name of 
sierra (saw) (Fig. 301). The topographic development of a range thus 
proceeds from youth into early maturity, and as erosion continues and 
the valleys widen, the declivities lessen, angularities of form tend to 
disappear, and the mass becomes more and more mature. The topo- 
graphic forms of the peaks and ridges and of the intervening valleys 
must depend largely on the nature and structure of the rock masses 
presented to erosion. 

The Jura Mountains of Switzerland (Fig. 274) present a type of some- 
what youttful dissection; here the folds themselves are the dominant 
topographic features, which erosion as yet has been unable to modify 
greatly. Many of the ranges in the Alps, the Himalayas, the Caucasus 
and the Rockies are in mature stages of dissection, and their folded 
strata have been breached in the anticlines and largely etched away. 
The terms youthful and mature in this connection are merely relative, and 
do not refer to absolute time; actually one range may be much older than 


another, and yet on account of its greater mass, difference in material, or 
difference in climate, be in an earlier stage of its life history. 

Later Stages of Erosion. If erosive processes continue their work of 
degrading a mountain mass, unhampered by further uplifts, the range 
gradually passes into a mature stage. The sharp peaks and asperities 
tend to disappear, the valleys to widen. The progress of the work goes 
more slowly as the slopes lessen, and as the resistant metamorphic 
and crystalline rocks of the inner core are reached. Thus a maturely 
dissected range presents rather smoothly rounded forms and outlines 
(Fig. 302), which contrast sharply with the angular features of the 
sierra type. As they wear down more and more and pass gradually into 
old age, we find these mountains in humid climates composed of massive 
rocks, of schists, gneisses, and granites, rather than of the limestones, 
sandstones, shales, and lavas of ranges in the earlier stages. In arid 
climates, on the other hand, limestones outlast the granitic rocks because 

Fig. 302. Characteristic forms and outlines of late-mature mountains. 

they are far less vulnerable to the mechanical weathering that character- 
izes dry regions. 

There appears to be no well-recognized term equivalent to sierra 
for mountains in these later stages; they are variously termed mature, 
subdued, or- old mountains. Examples are to be seen in the mountain 
masses of New England and eastern Canada, such as the Green Moun- 
tains, the White Mountains, and the Laurentian Mountains of Quebec; 
in Europe the Black Forest region of Germany and the Highlands of 
Scotland and Norway are examples. 

Final Stage : Peneplanation. Ultimately, provided no new upwarp- 
ing movements occur, the mountains will disappear and the region they 
occupied will be reduced nearly to baselevel. Since, however, the pro- 
cess of erosion goes on more and more slowly as the slopes lessen, it 
would evidently require an enormous lapse of time to bring down actu- 
ally to baselevel a mountainous tract, and we have no proof that this has 
ever in fact occurred. But we know that some areas have been reduced 
to low, almost featureless country; in other words to peneplains. Such 
country may still be diversified by scattered monadnocks projecting 
above the general level which, on account of their more resistant com- 
position, or possibly because of their position far from large streams, have 
not been reduced like their neighbors (Fig. 303). 

Disregarding occasional monadnocks, we may say that when the 


peneplain stage is reached the mountains have been obliterated; but we 
may yet be able to infer their former existence by the upturned and dis- 
located nature of the transversely eroded strata, by the widespread 
metamorphism of the rocks, by the slaty cleavage and faults which cut 
them, and by the presence of large granitic intrusive masses. We can 
not determine the former elevations, for, as LeConte has said, "we find 
only the bones of the extinct mountains 77 ; but from these remains we may 
learn the trend and extent of the ancient ranges. So, from the attitude 
of the rocks of southern New England, which is now only a hilly country, 
we are led to infer that it was once a mountainous region. 

Fig. 303. Stone Mountain, De Kalb Co., Georgia. A monadnock composed of granite 
which rises above the surrounding plain of erosion. (Geol. Surv. of Georgia.) 

Reelevation ; Complexity of Mountain History. If mountains were 
forever extinguished by the peneplanation at the end of a cycle, their 
history would be comparatively simple. But though the surface of the 
land may be smoothed out by erosion, the structure below the surface 
remains. The tilted and folded strata of unequal hardness and the ig- 
neous intrusions injected into them are still there, having disappeared as 
relief features because of peneplanation, but needing only a. second uplift 
of the land to bring them again into prominence. No new deformation 
of the rocks is required. A simple upwarping of the beveled surface/ 
such as has occurred again and again throughout the Earth's history, 
accomplishes the result. The sluggish streams are rejuvenated; they 
begin to cut actively, and they cut the weaker rocks most rapidly, leaving 
the resistant masses once more projecting above the surface. The moun- 


tains are thus etched again into relief, their new height depending en- 
tirely upon the amount and rate of the new uplift. 

Here is a range "in its second cycle/ 7 etched by streams from an up- 
lifted peneplain. How is the geologist to recognize the twofold nature of 
its history? The one indicator that is present in all mountains during 
the earlier stages of their second cycle is the close accordance of their 
summit levels, representing the surface of the former peneplain (Fig. 288). 
These summits will not endure throughout the second cycle, but they 
will be the last to be brought down toward the new baselevel because 
they are made of the most resistant rocks. 

Many ranges, however, have passed through an even more complicated 
history. Not a few have suffered uplift and denudation repeatedly. 
If each period of denudation had resulted in a peneplain, it would be 
impossible to unravel the Complete chain of events in such an intricate 
history; but wherever renewed uplift began to affect a mass well before 
the end of a cycle of erosion, some of the summits etched out during the 
earlier cycle were spared by the later, and thus old summits and broad 
rock terraces (page 418) were left as witnesses to what had happened. 

History of the Appalachians. Through close adherence to these 
principles, it has been possible to reconstruct the history of the Appala- 



Fig. 304. Ideal section across a part of the northern folded Appalachian ranges. 
The main stream is either superimposed or antecedent, trenching the hard ridges through 
seven water gaps (WrG). A former stream course, long since abandoned, is indicated 
by a wind gap (WdG). The present tributaries are subsequents (SS). Four successive 
baselevels (I, II, III, IV] are indicated by four accordant series of summit levels and 
valley floors, thus indicating four regional uplifts relative to sea level. Compare Fig. 288. 

chians. As we see them today they consist chiefly of long parallel ridges 
formed by the outcrops of very resistant rocks such as sandstones and 
conglomerates in a strongly folded sedimentary series (Fig. 304). Most 
of the ridges, even though narrow, have remarkably level summits, broken 
only occasionally by water gaps (Wrff) and wind gaps (WdG). Even a 
casual observer might notice that groups of these ridges reach a common 
level (II, Fig. 304), and if he examined them more closely he could see 


that certain of them reached a notably uniform higher level (7, Fig. 304). 
And if our observer became really interested and gave some attention to 
the broad valleys (III, Fig. 304) between the ridges he would discover 
what appear to be old valley floors deeply dissected by small streams into 
a network of low hills. The small streams drain into larger meandering 
subsequents (SSS, Fig. 304), flowing down the valley axes ; and all the 
meanders are entrenched. The subsequents in turn are tributary, in the 
northern and central Appalachians, to great streams which flow eastward 
indiscriminately across hard-rock ridges and soft-rock valleys. These 
are represented by the Delaware, the Susquehanna, the Potomac, the 
James, and the Roanoke. 

With these facts in mind, how much can we reconstruct of the Appala- 
chians' history? First, a period of strong deformation, as revealed by the 
great folds whose eroded stumps appear in the ridges. Second, erosion 
which must have begun during the first uplift, and have continued 
throughout a long cycle, resulting at length in a peneplain. This we 
can tell because the hard rocks in the folded series have been beveled 
down to a common level, as in the ridges I and in many others like them. 
Of course these rocks did not exist as ridges during the peneplain stage, 
but were merely parts of a low, gently undulating plain near sea level. 
Third, slow bodily uplift of the whole region, with rejuvenation, etching 
out of the resistant rocks into ridges again, and eventual reduction of the 
whole mass into a second peneplain at level II. The most resistant or 
most favored ridges (I) were not reduced, but remained as long narrow 
monadnocks. Fourth, a second slow upwarping, with the inevitable 
reetching of hard ridges from the peneplain II, and the development of a 
new baseleveled surface III on the soft rocks first attacked. This new 
surface is scarcely a true peneplain because of the great quantity of hard 
rock (all the I and II ridges) still unreduced. The meandering streams 
were just beginning to destroy these ridges when the third uplift occurred 
and forced them again to cut straight downward. Fifth, a third uplift of 
the land, rejuvenating the streams and thus causing them to entrench 
their meanders and with the aid of their tributaries to dissect the old 
baseleveled valleys 771 into the network of hills now existing. 

This, stripped of complicating minor movements and events, is the 
accepted history of the Appalachians; and if streams have behaved in the 
past as we see them behaving today, this history must be true. Author- 
ities are not as yet agreed as to the exact time at which each peneplain 
was formed, but the whole sequence of events from the orogenic period 
down to the present may have required 200 million years. 



Climatic Control of Denudation. The stages of the cycle of stream 
erosion under arid conditions have already been outlined (page 85). 
Complications of structure of course introduce changes in the progress of 
erosion and deposition. The effects, however, are less pronounced in an 
arid climate because of the shortness and rapid disappearance of the 
streams and because of the masking effect of the predominant alluvial 
fans. The important consideration here is that changes of climate bring 
about great changes in the landscape developed on any given set of 
structural conditions. Thus a change from a moist climate to a dry in a 
mountainous region causes the development of bolson basins; or an 
equally great change from a temperate climate to a cold may freeze the 
water into perennial ice and thus bring on glaciation with its resulting 
characteristic landscape. 

Structural and Lithologic Control of Denudation. The structure and 
lithology of a land mass are of vital importance to the landscape resulting 
from any given dynamic process such as stream erosion. The landscape 
developed in a mass of weak rocks elevated only slightly above the sea 
can never be more than one of rolling monotony even in maturity when 
relief is greatest. Rapid weathering and corrasion will provide an 
abundance of waste, the deposition of which, combined with lateral 
planation (much downcutting being impossible) results in the ready 
development of wide, open valleys. Similar streams will be able to act 
very differently, however, on a mass of resistant rocks elevated high 
above sea level. Steep gradients permit the development of scenic 
gorges whose sidewalls are accentuated in their steepness by the resist- 
ance of the hard rock to weathering and slope wash. Moreover, the slow 
inevitable changes wrought by erosion during the progress of a cycle in 
any land mass and under any climate, steadily alter the surface until the 
change has become profound. Thus it appears that the seemingly un- 
ending variety of landscape is in reality controlled by a few simple fac- 
,tors, and that slight variations in these factors produce the differences in 
scenery which lend enjoyment to travel and add to the richness of 
human existence. 


1. The Rivers and Valleys of Pennsylvania; by W. M. Davis. Geographical 
Essays, Boston, 1909, pp. 413-484. 

2. The Seine, the Meuse, and the Moselle; by W. M. Davis. Geographical 
Essays, Boston, 1909, pp. 587-616. * 

3. Earth Sculpture; by James Geikie. 320 pages. John Murray, London, 1898. 

4. The Scientific Study of Scenery; by J. E. Marr. Chaps. 8, 9, 10. Methuen 
& Co., London, 6th edition, 1920. 


Man wrests from the Earth many materials of the mineral kingdom 
for his necessities of life and comfort. The search for them has given 
rise to romance and adventure; their discovery has resulted in the open- 
ing up and settlement of new countries; their ownership has resulted in 
national, political, and commercial supremacy or has caused strife and 
war. Their richness has often been the incentive for man's acquisitive- 
ness. In the quest for these substances it is necessary to know about 
their distribution, occurrence, character, and origin, all of which is a part 
of the science of economic geology. Economic geology deals also with 
problems of investigation and other applications of geology to the uses of 
man, but we are concerned here merely with that part relating to mineral 

Of the great variety of mineral substances won from the Earth, coal is 
the most valuable, followed by the metallic minerals, petroleum and 
natural gas, and the nonmetallic minerals such as salt and feldspar. 
But since coal and petroleum are treated in Part II of this book, and 
many of the nonmetallic minerals are considered briefly in the different 
chapters, we shall restrict ourselves to the important group of metallic 
mineral deposits or, as they are commonly called, ore deposits. 

Ore deposits are geologic bodies that may be worked commercially for 
one or more metals. They are exceptional features, sparsely scattered 
in the rocks or on the surface; they constitute only an infinitesimal part 
of the Earth's crust, but they assume an importance far in excess of their 
relative volume because of the highly valuable materials they supply to 
natural wealth and industry. They have been concentrated in the rocks 
under peculiar and exceptional conditions that it will be our purpose to 
study. They cannot properly be considered apart from their geologic 
environment; consequently the information contained in the preceding 
chapters is vital to an understanding of them. A knowledge also of the 
origin and character of an ore deposit may aid in prophesying its size and 
depth beneath the outcrop. 


An ore deposit is of value for the metal or metals it contains, and these 
are usually locked up in one or more ore minerals. The latter in turn are 


commonly admixed with gangue minerals, and the mixture, which con- 
stitutes the ore is inclosed in the country rock. The term ore is often 
loosely used to designate anything that is mined from the Earth; but in a 
technical sense it denotes that part of a geologic body from which the 
metal or metals it contains may be extracted profitably. Nonmetallic 
substances such as coal, salt, feldspar, or building stone, which are used 
practically in the form in which they are extracted from the Earth, are 
thus excluded. A lead ore deposit, for example, may be inclosed in 
limestone country rock; the lead is chemically combined with sulphur in 
the ore mineral galena, and the latter may be admixed with the gangue 
mineral quartz, to form the ore. The winning of the lead from the ore 
deposit involves first a knowledge of the occurrence, shape, continuity, 
content, origin, and other geologic features of the deposit. Thus an 
understanding of the geology of the deposit is usually prerequisite for 
intelligent mining operations. The study of ore deposits, it will be seen, 
is preliminary to the other steps in the winning of metals. 

Ore Minerals. An ore mineral is one that may be used to obtain 
one or more metals. Thus galena is an ore mineral because it is mined 
for its metallic lead; but feldspar, although containing as much as 15 per 
cent of aluminum, is not an ore mineral because it is not mined for its 
aluminum content. The ore minerals occur as native metals or as chem- 
ical combinations of the metals with other elements. Gold and platinum 
usually occur as the native metals, and silver and copper are often found 
in that state. Most of the common metals are chemically combined 
with sulphur, arsenic, carbon, oxygen, or silica. 

Some ore minerals contain two or more metals, as, for example, 
chalcopyrite with its copper and iron, and individual metals may enter 
combinations to form several different ore minerals. In addition, several 
metals as, for example, silver, lead, and zinc, may occur in one deposit. 
Thus it is clear that the ore minerals of a deposit may represent a complex 
mixture of several metals, occurring in several different combinations, 
and with a single metal in more than one combination. In certain ore 
deposits, however, such as those of iron, this is not the case; only the one 
metal, iron, is obtained, and this occurs in just the one combination 
iron oxide. 

The metals of commerce are derived from many metallic combinations. 
Most of the world's gold has come from the native metal; consequently 
its removal from ores is a relatively simple process and offered no serious 
problem of extraction even to the ancients. Silver, on the other hand, 
is derived not only from the native metal but also from its combination 
with sulphur (sulphide). This is also true of the copper of commerce. 
Lead and zinc, however, are obtained chiefly from minerals containing 



sulphur, although combinations with carbon and oxygen contribute an 
appreciable amount. The vast quantity of iron used in industry is 
obtained almost entirely from combinations with oxygen (oxides). 
The simple metallic combinations just enumerated have yielded pure 
metals readily to the art of extraction and have supplied the human race 
with metals for over 2000 years. It must not be overlooked, however, 
that other less important and more complex metallic combinations not 
mentioned above yield appreciable amounts of the common metals as 
well as many of the minor metals not considered in this chapter. 

Some of the important ore minerals from which the common metals 
are extracted are listed below, and several of these are described briefly 
in Appendix A. 

List of the Commoner Ore Materials 


Ore Mineral 


Percentage of Metal 


Native gold 




Native silver . 




Silver, sulphur 


Native copper. . . . 




Copper, iron, sulphur . 




Copper, iron, sulphur 


Chalcocite . . 

Copper, sulphur 



Copper oxygen 


Malachite . 

Copper, carbon, oxygen, water 



Copper, carbon, oxygen, water 



Lead sulphur . . . 




Lead carbon oxygen 



Lead, sulphur, oxygen 



Zinc sulphur 




Zinc carbon oxvgen . . 



Zinc silica 



Zinc, oxygen 



Iron oxygen . . . .' 




Iron, oxygen 



Iron oxygen water, 



Iron carbon, oxygen 


Gangue Minerals. Gangue minerals are the valueless minerals of 
the ore, and are usually earthy or nonmetallic in character. In common 
usage they are simply referred to as gangue an old mining term. Thus 
in an ore deposit containing quartz and galena, quartz is the gangue. 
Several gangue minerals may be present in one deposit. Some of the 
common gangue minerals are quartz, calcite, dolomite, siderite, and 


Gangue minerals are lacking in some deposits, and then the inclosing 
country rock is sometimes loosely referred to as gangue, so the term is 
somewhat flexible. Some of the gangue minerals considered worthless 
today may under improved metallurgical processes turn out to be valu- 
able ore minerals of, tomorrow. 

The Ore. It will be evident from the foregoing statements that ores 
vary greatly in mineral content and chemical composition. No two 
are ever exactly alike. There are simple ones iron, for example, 
which contain no gangue minerals and are composed solely of the ore 
mineral hematite. This is commonly thought to be the case with the 
ores of other metals, but it is far from correct. The ore minerals usually 
constitute but an insignificant part of the ore. If one were to take a 
trip through a profitable gold mine he might search the ore in vain to see 
a speck of gold; gangue minerals alone would meet the eye, since the 
weight of gold may form only a few ten-thousandths of one per cent of 
the ore. The proportion of ore to gangue minerals in deposits of copper, 
lead, and zinc, however, falls between the extremes of iron and gold, but 
the gangue usually predominates. , 

The abundance of metals in ores is also reflected in their relative prices, 
for a ton of iron may be purchased at about the same price as an ounce of 
gold. The higher the value of the metal the lower the grade of ore that 
can be mined for the same cost. Thus, metals such as gold or platinum 
make profitable ore with only a few tenths of an ounce of metal in each 
ton of ore. It is evident that the tenor or metallic content of ore that 
may be profitably mined depends largely upon the selling price of the 
different metals, the size and character of the deposits, and their accessi- 
bility. Thus economic as well as geologic factors determine what 
constitutes ore. The tenor obviously varies with deposits of different 
metals and with different deposits of the same metal. Man, of course, 
imposes no upper economic limit to the tenor of the ore; the richer the 
better. But the lower limit is vital, since sufficient metal must be ex- 
tracted to pay for the cost of producing it and to yield a profit. One 
might own a deposit containing gold, but if each ton of rock contains 
only $1.00 of gold and it costs $2.00 to extract that amount of gold from 
the ton of rock, then obviously the deposit has no value. But what is 
not ore today may with improved processes of extraction and transporta- 
tion b ore tomorrow. 

Most of the gold produced today comes from ore that contains from 
0.1 to 0.3 ounce, or $2.00 to $6.00 worth of gold per ton of ore, but small 
deposits containing $10.00 per ton cannot be mined profitably in certain 

Silver ores range from 5 to 25 ounces of silver per ton of ore, whereas 


most of the copper of the world is obtained from large ore deposits that 
have less than 40 pounds of copper per ton of ore. Zinc ore must con- 
tain from 3 to 30 per cent of zinc; and lead ore from 2 to 10 per cent 
of lead. Iron ore from the Lake Superior region contains 40 to 60 per 
cent of iron. 

Gold and silver are commonly associated with the other metals, and 
their presence may enable ore of lower grade than the figures given 
above to be worked. More than one metal may be won from certain 
ores; thus lead and zinc, copper and zinc, and silver and lead are common 


When one considers that ore deposits represent concentrations of 
unusual minerals and that they are sparsely scattered in the Earth's 
crust, immediately the questions arise: Where did the metals come from, 
and how did they become concentrated into relatively small and widely 
scattered bodies; in short, how were the ores formed? 

The ore deposits as we see them today occur in diverse forms in all 
kinds of rocks under conditions which preclude the possibility that they 
owe their origin to the operation of any one process. We shall first 
consider their source and next the means by which they have been col- 
lected, carried, and deposited in the forms in which we now find them. 

Source of the Metals. The metals must have come from within the 
Earth, where the igneous rocks originate. They are so widely and 
intimately associated with igneous rocks or other indications of igneous 
activity that the two must have had a common origin. Although traces 
of the metals occur in sedimentary rocks, the sediments themselves were 
originally derived from igneous rocks, and their metal content likewise is 
most likely of igneous origin. Of course it must be remembered that 
ores, like rocks, may be eroded or dissolved and carried elsewhere, and 
their materials may be formed again in other places. Thus some of 
them may have obscure parentage. 

The conclusion that ores and igneous rocks originate together is 
further substantiated by the occurrence of traces of most of the metals in 
the'igneous rocks and in hot springs or other emanations of igneous origin. 

Collection and Transportation. The above conclusion of the ulti- 
mate source of the metals might lead one to infer that all ore deposits 
must occur in igneous rocks; but this inference is not correct. Some 
deposits, it is true, evidently are an original part of an igneous mass, but 
others fill cracks in igneous rocks and therefore must have been formed 
after the igneous mass solidified. Moreover, many lie in sedimentary 
rocks far distant from an igneous mass. Consequently we conclude* 


that in some cases the metals have been collected within a magma cham- 
ber and have remained in the intrusive when it solidified, and in other 
cases they have been gathered up by some mobile agent, expelled from 
the magma chamber, and carried some distance to places where deposi- 
tion of the metals occurred. Thus, some ore deposits are component 
parts of igneous masses and have been formed at their source by solidi- 
fication from a magma state, whereas others have been formed as a result 
of deposition from mobile carriers. 

To the first of these we give the name igneous or magmatic deposits. 
The substances of which they are composed are more or less common to 
certain kinds of igneous rocks, and presumably were originally diffused as 
minute particles of metals or metallic compounds throughout the mother 
magma. They 'were collected and concentrated during the cooling of the 
magma by the process of magmatic differentiation (page 438), and they 
solidified or crystallized more or less simultaneously with, and as a part 
of, the original igneous body. Such deposits are, therefore, simply 
unusual kinds of igneous rock that happen to be of value because their 
ingredients are desired by man; otherwise they would be looked upon as 
varieties of rock. Their mode of origin, therefore, is the reason for their 
name. It must not be supposed, however, that all magmas during their 
crystallization give rise to these ore deposits. If this were the case such 
deposits would be common features, whereas they are rare. 

The character of the deposits formed in the manner described above 
will be given attention later. Next, we shall consider the processes by 
which the metals have been collected and transported from the magma. 

The conclusion is inescapable that the carriers of the metals must have 
been in gaseous or liquid form, depending upon the state of consolidation 
of the magma at the time they were expelled. This conclusion loses 
some of its strangeness if we stop to reflect upon the information available 
regarding such substances. 

Highly heated gases are well known to be a part of magmas. In fact, 
copious exhalations of hot gases take place during volcanic eruptions. 
Moreover, it is known that they carry metals, since tests made upon 
such volcanic gases as those of the Valley of Ten Thousand Smokes, and 
of other places, prove this conclusively. There are also many other 
observations that lend support to the above statements, but they are too 
detailed to be considered here. We conclude, therefore, that such highly 
heated gases have been one important factor in collecting and transport- 
ing metals to their present resting places. 

Liquids are believed to have been an even more important agent in 
transporting metals. It is generally thought today that by far the great- 
est number of ore deposits have been formed by means of hot waters. 


Their association with igneous rocks is well known. They are abun- 
dantly emitted from volcanoes, and seepages of hot water in the form of 
hot springs continue in volcanic regions long after eruptions have be- 
come quiescent. Furthermore, the rocks adjacent to deep-seated igne- 
ous intrusives give evidence that they have been traversed by hot 
waters that came from the cooling igneous bodies. There is no 
question, then, that hot waters are normal emissions from cooling igne- 
ous bodies. 

That hot waters can and do dissolve metals has been demonstrated 
over and over again in the chemical laboratory. But if doubt remains 
that natural hot waters are competent solvents and carriers of most of 
the metals, it is dispelled by the array of tests made upon the hot spring 
waters of Steamboat Springs, Nevada, and on those of other localities, 
which show the presence of dissolved metals. Metallic minerals are 
actually being deposited from some such hot springs. Hot waters are 
also occasionally encountered in deep mines, and they too contain metals 
in solution. Hot magmatic waters, therefore, are also carriers of metals 
from their magma source. 

Hot waters follow the gaseous emissions and represent a later and cooler 
phase of igneous activity. Both gases and waters obtain their load of 
metals from a cooling magma. 

There is still another mode of collection and transportation of metals 
by liquids. Ordinary meteoric waters, that is, those originally derived 
from rain, play an important part in the formation and alteration of ore 
deposits. The commonest example is that of cold surface waters, which 
dissolve metallic compounds from the upper parts of ore deposits and 
carry them down beneath the ground water table. Also in the weather- 
ing of rocks certain metals, such as iron, are dissolved and carried along 
by surface waters to places where deposition of their metallic content 
later takes place. Artesian waters are believed to have dissolved and 
carried great quantities of lead and zinc, which later were deposited to 
form the extensive ore deposits of the Mississippi Valley. It is thought, 
too, that the waters of many hot springs are meteoric waters that have 
moved downward into hotter regions and have risen again to the surface 
as hot waters. Such waters also collect and transport metals. 

Thus it is seen that as steps in the origin of ore deposits, metals have 
been concentrated in the magma and deposited by solidification within 
its crystallized mass; they have been collected from their magma source 
by highly heated gases and hot waters and carried to the outer cooled 
portion of the intrusive or into distant rocks; they have been searched out 
of cold surface rocks by meteoric waters and transported elsewhere. 

A further step is necessary, however, to bring about the formation of 


ore deposits the metals must be deposited from their carriers; and we 
shall now consider the means by which this is accomplished. 

Deposition. It is obvious that the manner of release of the metals 
from their carriers depends somewhat upon the nature of the carrier. 
Thus different processes operate to bring about deposition from vapors 
and gases than from liquids. 

Metals are given up from vapors in two ways: (1) by a decrease in 
temperature and pressure due to contact with cooler rocks; this lowers 
the solvent power of the vapors and necessitates deposition of minerals; 
(2) by a chemical reaction between the vapors and the rocks with which 
they come in contact. This process has already been considered under 
the heading of Contact Metamorphism (page 354), where it was shown 
that the hot gases given off by a cooling magma produce profound min- 
eral changes in the adjacent country rock. If the gases contain also 
notable quantities of metals, contad-metamorphic deposits are formed as 
an incidental phase of the contact metamorphism. The nature of these 
deposits will be described later. 

Deposition from hot waters takes place with or without chemical 
reaction with the wall rocks, and different kinds of ore deposits result 
from each process. In the case of chemical reaction, the hot solutions 
diffuse through or insinuate themselves along cracks or other openings 
in the rocks, and dissolve all or part of the rock, particle by particle. 
Simultaneously they deposit equivalent volumes of ore and gangue 
minerals. Ore replaces country rock, and for this reason the resulting 
deposits are called replacement deposits. The process may be roughly 
illustrated by supposing that the clay bricks of a wall could be replaced, 
one by one, by silver bricks; the resulting silver wall would have the same 
position, volume, and structure as the replaced brick wall. In nature, 
however, the particles, instead of being the size of bricks, are of molecular 
size, and the substitution takes place by chemical action in a solution. 
The substitution may start at a number of centers and give rise to dis- 
seminated deposits. The centers may enlarge until they coalesce and a 
large volume of rock is thus more or less completely replaced, forming 
massive replacement deposits. If only the walls of fissures are replaced, 
replacement veins are formed. This replacement process may go on until 
large ore deposits result. The chemical reaction between gases and rocks 
mentioned above is also replacement. 

If reaction with the wall rocks does not take place, the metals stay in 
solution until other processes bring about their deposition. In their 
travel through the rocks they penetrate fissures, joints, pore spaces, 
caves, or other rock openings. In such places opportunity is afforded 
for deposition to take place; the cavities become filled with ore and the 


resulting deposits are called cavity-filled deposits. Deposition in these 
openings may be brought about by several factors. An important one 
is the lowering of the temperature of the solutions by their passage 
through the rocks, making the substances carried less soluble, with the 
result that they are precipitated as minerals. Or, there may be changes 
in the concentration of the substances in solution, or the solutions may 
react chemically with other solutions, gases, or minerals, to bring about 

Deposition of metals from meteoric waters may take place by some of 
the processes mentioned above, or by evaporation such as takes place 
when sea water is evaporated and salts are deposited. Organic materials, 
such as plants and certain forms of bacteria, are also thought to be ef- 
fective agents in depositing iron from surface waters. 

It is thus evident that ore deposits do not owe their origin to any one 
simple process, and various types of deposits are formed as a result 
of the operations of the different processes outlined above. Before we 
take up examples of the different types of deposits, it is desirable to 
consider briefly some other physical factors that influence ore de- 

Effect of Temperature and Pressure upon Ore Deposition. Tem- 
perature and pressure play an important part in the character and loca- 
tion of ore deposition. This is particularly true of deposits formed from 
igneous emanations. During the earlier stages of an intrusion the 
emanations consist of highly heated gases and vapors, as no liquids can 
exist at the temperatures that prevail. As they move toward the surface 
they pass through zones of decreasing temperature and pressure. 
Gradually they change to hot waters, and deposition of different ores 
results in response to the changing physical conditions. Certain miner- 
als, such as pyrite, form under a wide range of temperature and pressure, 
but the deposition of others is restricted to definite ranges of temperature 
and pressure; consequently they become diagnostic of those particular 
conditions. Thus, those deposits formed by hot gases and vapors near 
the contact of the intrusive are characterized by certain associations of 
minerals of high-temperature origin, such as make up contact-metamor- 
phic deposits (page 439). As the emanations move farther toward the 
surface, they form deposits which have been divided by Lindgren into 
three classes: 

1. High-Temperature Deposits. These are formed at great depth and 
under high temperature by filling fissures or replacing the country rock. 
They are characterized by such high-temperature minerals as garnet, 
pyroxene, amphibole, and magnetite, and contain gold, tin, iron, and 


2. Intermediate Deposits. These also fill openings, or replace the 
country rock. High-temperature minerals are absent, and such min- 
erals as quartz, calcite, pyrite, chalcopyrite, galena, and sphalerite are 
typical of them. The deposits furnish gold, silver, copper, lead, and zinc. 

3. Low-Temperature Deposits. These have been formed at lower 
temperatures and shallow depth, and occur chiefly in fissures in shattered 
rock. They supply most of the gold, silver, and mercury of the world, 
and are characterized by such minerals as quartz, chalcedony, carbon- 
ates, and gold and silver minerals. High-temperature minerals are 
lacking. Most of the gold-silver deposits of the Rocky Mountain 
States belong to this group. 

Thus it is evident that the environment controls in large part the char- 
acter of the ore deposits that are found, and the type of deposit found on 
the present surface depends largely upon the depth to which erosion has 


From the foregoing section it is evident that ore deposits are not all 
alike nor are they' simple geologic bodies. They have been formed by 
different processes, are composed of numerous substances, and occur in 
many forms. Innumerable kinds of ore deposits are the result. A few 
of the more important types will now be considered, 


Igneous Ore Deposits. Since these deposits originate by solidifica- 
tion from magmas, they usually occur in or near intrusive igneous rocks. 
Their shapes are irregular and they vary greatly in size. They yield 
magnetic iron ores, corundum, chromium, platinum, nickel, and copper. 
The deposits seldom consist of masses of pure ore minerals; varying 
amounts of rock minerals are mixed with them. Bodies of magnetite of 
course cannot contain much admixed rock mineral, or they will have no 
value as iron deposits. On the other hand, platinum need constitute 
only a fraction of one per cent of the rock to make a valuable deposit. 

Some ores are usually associated with certain magmas; for example, 
nickel-copper deposits occur only with a variety of gabbro, and chromium 
and platinum with peridotite and allied rocks. An area of peridotite is 
always worthy of search in the hope that chromium or platinum deposits 
may be discovered in it. 

Examples of igneous deposits occur the world over. The immensely 
valuable magnetite deposits at Kiiruna, Sweden, where 750 million tons 
of iron ore exist, are of this type, as are also many of the magnetite 
deposits of the Adirondack region in New York State. The richest 
nickel deposits of the world, at Sudbury, Ontario, are igneous deposits. 



The great platinum and chromium deposits recently discovered in South 
Africa are also igneous deposits. 

Contact-Metamorphic Deposits. Certain conditions are necessary 
for the formation of this type of deposit. The intrusion must be deep- 
seated; the intrusive is usually feldspathic, such as granite or diorite, and 
valuable deposits occur only where the intruded rocks are limestones or 
limy shales. It does not follow that all intrusives exert contact meta- 
morphism or that all contact metamorphism is accompanied by ore 
deposition; in fact, ores are the exception rather than the rule. 

Contact-metamorphic deposits usually consist of several bodies of 
ore irregularly scattered throughout the contact-metamorphic aureole. 

Pig. 305. Diagram showing a cross section of contact-metamorphic deposits. 
Stippled area represents contact-metamorphic zone and black, ore. 

Most of them, however, lie either adjacent to or quite close to the in- 
trusive (Fig. 305). The ore bodies vary greatly in shape and size. The 
gangue consists of the metamorphosed rocks described in Chapter XIV, 
and such minerals as garnet, amphibole, and pyroxene. Magnetite, 
hematite, and the common sulphides of iron, copper, lead, and zinc are 
intimately admixed with the silicates. The ore minerals are commonly 
scattered in small particles throughout the gangue and usually constitute 
a minor part of the ore. Some of these deposits are worked for iron; 
others for copper, zinc, lead, gold, tungsten, or molybdenum. They are 
not as numerous as other types of ore deposits, though individual bodies 
may be of great value. Usually they are low-grade, and their irregularity 
of shape, size, and distribution makes them difficult and financially haz- 
ardous to mine. They occur in regions where extensive erosion has 
revealed the larger deep-seated intrusives, and examples are numerous in 
the eastern and western United States and in Scandinavia. Copper 
deposits of this type have been mined at Morenci, Arizona; gold deposits 
at Hedley, British Columbia; lead and zinc deposits at Hanover, New 
Mexico; iron deposits at Cornwall, Pa., and Banat, Hungary. 

Replacement Deposits. Replacement deposits occur in rocks of all 
kinds; but limestones are the most common hosts because they are more 
readily attacked by mineralizing solutions. Replacement deposits are 


usually irregular in shape and some of them attain great size (Fig. 306). 
Many of the world's largest bodies of ore belong to this class. Some 
deposits consist of disseminated ore minerals sparsely scattered through 
huge volumes of rock. Such low-grade deposits in the southwestern 
United States, where they are worked on a large scale, have made this 

country the world's greatest pro- 
ducer of copper. These are in 
sharp contrast with the massive 
deposits in which the replacement 
process has gone so far that prac- 
tically no original rock is left 
within the ore. 

The deposits are worked for 
I - S + T many metals; vast quantities of 

pended blocks of unreplaced limestone in the " . 

ore (dark) show that the rock was replaced by COpper, lead, and Z1BC, and COn- 

ore bit by bit, otherwise the inclusion would gjderable amounts of gold and sil- 

not be suspended. . 

ver are obtained irom them. 

Examples of replacement deposits are numerous throughout the 
western and southwestern States. The great silver-lead-zinc deposits of 
Leadville, Colorado, with a total output of 450 million dollars, are of this 
type. In the Coeur d'Alene district of Idaho silver-lead and zinc de- 
posits in quartzite have yielded over one-half billion dollars. Each ton 
of ore, as mined, contains about 8 per cent of lead (160 pounds) and 6 
ounces of silver. The large, massive replacement deposits of the Huelva 
district, in Spain, consist of pyrite with copper, inclosed in slates and 
porphyry. They have been worked since the time of the Phoenicians 
and have produced 175 million tons of pyrite with 3 to 4 million tons of 

Cavity-Filled Deposits. There are many kinds of cavities in the 
rocks, and any of them may become filled with ore, each giving rise to 
a deposit of different form. However a certain amount of replacement 
of the walls of cavities has taken place while the cavities themselves 
were being rilled by ore. Fissures are the most important class of 

A fissure vein in its simplest form is a fissure or fracture in the rock 
rilled with mineral matter deposited from solution (Fig. 307). The fis- 
sures may or may not be faults. Fissure veins are numerous, particu- 
larly in mountainous regions where rocks are folded or where igneous 
intrusions have occurred. Their origin by deposition from solutions is 
commonly shown by coatings or crusts of different minerals parallel to 
the walls which form sharp boundaries with the country rock (Fig. 308). 
The central layers may not meet, in which case there remain unfilled 


spaces, lined with projecting crystals, called vugs (Fig, 308). Many 
rare and beautiful crystals are formed in these vugs. 

There are a great many fissure veins that contain only worthless gangue 
minerals such as quartz; many also contain insufficient ore minerals to 
make ore. In fact, those that do contain good ore are relatively few; 

Fig. 307. Fissure vein of gold-quartz ; the mining discloses the width of the vein 
and the wall rock on either side. Cook Mine, Colorado. 

otherwise mines would be more numerous than they are. Variation in 
metallic content constitutes one of the uncertainties of mining operations. 
There may be good ore in one place along a vein and lean material in 
another. Those portions in which the ore minerals are more concen- 
trated are called ore shoots; a vein usually contains one or more of them. 
In addition to the actual vein filling, the walls of most fissure veins are 
impregnated with ore minerals. 

Fissure veins may be vertical or inclined, and occur singly, in parallel 
groups, or in intersecting groups. The intersections intrigue the miner 



in the hope of obtaining rich ore. Some fissure veins have been followed 
several thousand feet along the strike and as far down the dip. The 

majority, however, are relatively 
shallow and die out within 2500 
feet of the surface. In thickness 
they range from a few inches to 
tens of feet, but most of them are 
less than 10 feet. Fissure veins 
terminate along their strike and 
dip by pinching out or abutting 
against other fissures. Commonly 
I) b iTb ' they are cut and displaced by faults, 

Fig. 308. Section of a fissure vein; aa, and the finding of the faulted por- 

wall rock; bb crusts of ore minerals; cc, tiong j g Qne of the pro blems in min- 

gangue minerals; a, vugs lined with project- f . c 

ing crystals. Commonly the filling is sepa- ing Operations. 

rated from the wall rock by a thin layer of JTi ssure ve i ns have been enor- 
crushed rock known as gouge. 

mously productive of silver, gold, 

copper, mercury, antimony, and other metals. They are world-wide in 
their occurrence. The veins of Butte, Montana, have yielded from 1882 
to the end of 1928 two billion dollars in copper, silver, gold, and zinc; those 
of California over 600 million dollars in gold; and those of Western 
Australia over 800 million dollars in gold. A single vein, the Comstock 

Fig. 309. Cavity-filled deposit. Shows an open cavity coated by deposited layers 
of gangue and ore mineral, the latter black. The stalactites and stalagmites of the origi- 
nal opening show that the cavity existed before the deposition. Masses of ore mineral 
fell from the roof before the last deposit was made. 

Lode of Nevada, has contributed nearly 400 million dollars in gold and 
silver; mercury has been produced from the veins of Almaden, Spain, for 
over 2000 years. 

Other types of cavity-filled deposits have been formed by the filling of 
caves, joints, bedding planes, cleavage planes, breccia openings, and rock 
pores. The first three have given rise to the productive lead and zinc de- 
posits (Fig. 309) in Wisconsin. The filling of pore spaces in the vesicular 


basalts and interbedded conglomerate of the Lake Superior copper dis- 
trict accounts in large part for the 4 million tons of metallic copper that 
have been mined from this famous district down to depths of nearly 6000 
feet beneath the surface. 

Sedimentary Ore Deposits. In places where sediments are accum- 
ulating, certain beds whose component minerals happen to be of com- 
mercial value may be deposited from solutions to form ore deposits. 
Sedimentary iron or manganese ores are formed in this manner (Part II, 
pp. 164 and 278). Sedimentary iron deposits may be as continuous and 
as extensive as the sedimentary series in which they occur, and obviously 
cannot extend beyond its limits. The sedimentary ore beds manifestly 
partake of the structure of the rest of the sedimentary rocks that inclose 
them, and may be horizontal, folded, or faulted. Consequently their 
shape and distribution can be accurately determined by a study of the 
structure of the region in which they occur. 

The Clinton hematite beds, with their 600 million tons of available 
iron ore, are the most noteworthy representatives of sedimentary ore 
deposits in North America. They extend from New York State to 
Alabama and support the great steel industry of Birmingham, Alabama. 
They have an area of hundreds of square miles, though they are not 
thick enough to be worked in many places. The ore beds are hard red 
hematite and average 35 per cent of iron. The iron ores of Lorraine, in 
France, are another example. These beds are the most important in the 
world and are estimated to contain 5000 million tons of iron ore. 


We have already seen in Chapter III how the rocks become weathered 
and disintegrated by mechanical and chemical agencies. Ore minerals 
in the rocks are likewise affected. Some become broken up and are 
transported as detritus, others are taken into solution and thus carried 
away, and insoluble ones may remain behind while the materials sur- 
rounding them are dissolved and removed. If the valuable materials 
thus released from the rocks later become concentrated, different kinds of 
secondary or disintegration deposits result. 

Mechanical Concentrations : Placers. In this type of deposit 
nature has operated to produce the results achieved by man when he 
mines, crushes, and concentrates ore to obtain the desired ore minerals. 
The ore minerals concentrated in ore deposits or those that are sparsely 
scattered in the rocks become separated from the surrounding gangue or 
country rock by mechanical and chemical agencies. The disintegrated, 
materials move slowly down the surface slopes to the nearest stream, 


where the water sweeps away the lighter rock and gangue particles, and 
the heavier ore minerals sink to the bottom or are moved relatively short 
distances. As thousands of tons of debris are thus moved to the streams, 
the few ore minerals in each ton of debris will be concentrated in the 
gravels of the stream bottom until there is accumulated a deposit of 
sufficient size to be workable. These are called placer deposits, and the 
operation of extracting the valuable minerals is called placer mining in 
contrast to lode mining from bedrock deposits. 

Certain conditions are necessary for the formation of placer deposits; 
the ore minerals must be insoluble,- or they will be taken into solution; 
they must be heavy enough to sink in moving water that will sweep away 
the same sized particles of rock and gangue; the streams must have 
sufficient velocity to sweep away the rock and gangue particles. If the 
stream velocity is great the ore particles also will be swept along until a 
place is reached where the current slackens; there they will be dropped. 
If, for example, a quartz vein containing $1.00 of gold in each ton of ore 
is eroded, the gold particles, being insoluble and heavy, will concentrate 
in the stream bottom gravels while most of the quartz is swept away. 
Eventually the gold content of many thousands of tons of ore is thus 
concentrated. The rich deposits formed in this manner have given rise 
to the great California gold rush of 1849, the Klondike stampede to the 
Yukon, and the rich discoveries in Alaska, Australia, and other places. 
Hundreds of millions of dollars in gold have been extracted from placer 

Relatively few minerals meet the requirements outlined above, for 
most of them, as will be seen later, are chemically attacked by surface 
water, or are of low specific gravity. Gold is by far the most common 
placer mineral, but platinum, tinstone, magnetite, quicksilver, and 
precious stones also occur in placer deposits. 

The placer ore minerals are disseminated in gravels; occasionally 
pockets or streaks, called bonanzas, occur in which a shovelful of gravel 
may contain a hundred dollars or more in gold. Most placer gold is in 
the form of fine specks called "dust," but to the joy of the miner larger 
lumps or nuggets are also found, some of which have a value of several 
thousand dollars each. 

The earliest primitive mining undoubtedly was from deposits of this 
type. The ease of extraction and the richness of some deposits makes 
them eagerly sought. The hardy miner requires only a shovel and pan 
to extract the gold; a shovelful of gravel in a pan is dextrously rotated in 
water until the gravel is washed free from the gold. But this method 
suffices only for the richer deposits; low-grade gravels are worked by 
sluicing (Fig. 310); jets of water wash the gold-bearing gravels through. 


sluice boxes, and the gold collects on riffles or cross bars in the bottom. 
In ancient alluvial mining, fleeces were placed in the boxes instead of 
riffles; hence the origin of the fable of the Golden Fleece. More refined 
methods utilize large mechanically operated dredges that can handle 
profitably great volumes of gravel containing as little as 8 cents 7 worth of 
gold in each cubic yard of gravel. Gold gravels are worked in western 
North America, Alaska, Yukon, Australia, New Zealand, and Africa. 
Tin gravels are mined extensively in the East Indies, and most of our 

Fig. 310. Placer mining by sluicing. The gravel containing the specks of gold is 
washed by powerful jets of water into sluice-boxes, where the gold is caught by riffles. 
Cariboo District, British Columbia, Canada. 

platinum is won from placer gravels, particularly in the Ural Mountains, 

Placer deposits are not restricted to present stream bottoms; the 
shifting and downcutting of streams has left placer gravels stranded on 
hillsides and in stream terraces. In arid parts of Australia the wind has 
served, instead of water, to concentrate the gold, by blowing away the 
lighter rock matter. 

Residual Concentrations. Relatively insoluble minerals such as 
manganese, aluminum, or iron, occur scattered through soluble rocks 
and, in the process of weathering, the rock may be dissolved and carried 
away, leaving the insoluble particles to accumulate. The process, if 
long continued, gives rise to residual accumulations of iron, aluminum, 



or manganese ore. Similarly, insoluble gold contained in soluble pyrite 
undergoes residual accumulation. 

Chemical Concentrations. in the process of weathering and erosion 
some ore minerals are taken into solution, transported short distances, 
and redeposited elsewhere on or near the surface. Many deposits of 
bog-iron ore are formed in this manner. Other minerals as well are dis- 
solved and carried down into the original ore deposit and there precip- 
itated, but as this is involved with other changes that affect the upper 
parts of ore deposits, it will be considered under the following heading. 


When bedrock ore deposits become exposed at the surface by erosion, 
they are weathered along with the inclosing rocks. The surface waters 

and their contained gases oxidize 
many ore minerals and yield sol- 
vents which dissolve other miner- 
als. Thus the upper part of an 
ore deposit becomes oxidized and 
leached down to the water table, 
and the part thus weathered is 
called the zone of oxidation. As 
the cold, dilute leaching solutions 
trickle downward through the de- 
posit, they may lose a part of their 
metallic content in the zone of oxi- 
dation, but when they reach the 
ground water certain of the dis- 
solved metals are precipitated as 
sulphides to form the secondary en- 
Fig. sii. Diagram showing zonal ar- richment zone. The lower unaffect- 

rangement of a weathered vein: aa country ed part O f the deposit IS Called the; 
rock; 66, water table; cd, vein; d~f, oxidized . . ^ 

zone; d, capping or gossan; e, leached por- primary zone. This zonal arrange- 

tion; /, oxidized ores in oxidized zone; g t ment ^'lg. 311) is characteristic of 
secondary enrichment zone; h, primary zone. 111 

weathered ore deposits. In places 

the secondary enrichment zone may be absent, and rarely the oxidized 
zone is shallow or lacking, as in glaciated regions. The changes that 
take place in the different zones will now be considered in more detail 
and the processes will be illustrated by referring to a simple copper vein 
consisting of quartz, pyrite, and chalcopyrite. 

Changes in the Oxidized Zone. The familiar rusting of iron upon 
exposure to the weather takes place also in the pyrite of an ore deposit. 
It becomes chemically altered, and limonite, sulphuric acid and ferric 


sulphate, all products of oxidation, are formed. The ferric sulphate is a 
ready solvent for many ore minerals; it dissolves the copper in the chal- 
copyrite, forming copper sulphate, which slowly trickles down through 
the crevices in the deposit. The limonite, however, does not go into 
solution but remains behind and stains the quartz a rusty color, giving 
rise to a gossan or iron hat. ^ 

These rusty, cavernous gossans are the surface indications of mineral 
deposits but usually they do not disclose very much about the mineral 
character of the ore because the ore minerals have been removed by the 
leaching solutions. Gossans may result also from worthless pyrite, so 
that the mere presence of a gossan does not indicate a valuable mineral 
body. The value usually cannot be determined until at great expense 
the deposits are penetrated at depth. Certain features of gossans may 
indicate to the trained geologist something of the original mineral char- 
acter, but their discussion is beyond the scope of this work. As the 
oxidation and leaching extend downward usually to the water table, the 
depth of the leached barren zone may, in arid regions, reach several 
hundred feet. Because of this, valuable deposits have remained undis- 
covered for long periods of time. 

Silver, zinc, and other minerals also are leached from the oxidized zone. 
Gold, however, is rarely removed; it remains behind as native gold in the 
rusty quartz of the oxidized zone. Consequently gold deposits, unlike 
others, commonly are actually richer in the oxidized zone because they 
have undergone residual enrichment. For example, if primary ore with 
$5.00 of insoluble gold per ton is one-half removed by solution, there is 
then left $5.00 in gold in a half-ton, or $10.00 in a whole ton. The ore is 
thus doubled in value in the oxidized zone by residual enrichment, ,and 
contrary to the usual conception, becomes leaner in depth beneath the 
water table. 

The sulphate solutions of the metals formed in the oxidized zone in 
their journey down through the deposit may lose a part of their metallic 
content. The copper sulphate, for example, may undergo evaporation 
and green copper sulphate minerals will be deposited; it may react with 
calcium carbonate to form the blue (azurite) and green (malachite) 
carbonates of copper; or with soluble silica to form the light-blue copper 
silicate (chrysocolla). Native copper also may be deposited. Similarly 
zinc and lead minerals, native silver, and other minerals of metals are 
deposited in the oxidized zone. In places these minerals constitute large 
and valuable ore deposits, but they must always be expected to disappear 
beneath the zone of oxidation. 

Secondary Enrichment Zone. When the sulphate solutions reach the 
water level or a place where no oxygen is available, they undergo a 


chemical change that causes deposition in the form of secondary sulphide 
minerals such as covellite (CuS) or chalcocite (Cu 2 S). Thus, to use one 
metal as an illustration, copper has been taken out of the upper part of an 
ore deposit and added to a lower part, thereby enriching the lower part in 
copper and forming the secondary enrichment zone. This zone may 
extend downward several hundred feet but eventually it gradually merges 
into the unchanged primary zone below. This process of leaching above 
and deposition below may go on continuously as erosion lowers the 
water level, until primary material that contained only a half of one per 
cent of copper has been enriched to workable ore containing as much as 
3 or 4 per cent of copper. Most of the great copper production of the 
United States comes from ores that have been thus enriched. If it had 
not been for the operation of this process, such large mines as those at 
Bingham, Utah; Santa Rita, New Mexico; Ely, Nevada; Miami and Ray, 
Arizona; and others, would not exist. 

It is evident, therefore, that some weathered ore deposits are barren 
in the oxidized zone, rich in the secondary zone, and lean in the primary 
zone; or, in the case of residual enrichment, are richer above than below. 
The recognition of these surficial changes in ore deposits is one of the 
modern achievements in geology. It is of great practical importance 
because it has stimulated deep exploration beneath barren gossans, with 
the result that many valuable ore deposits have been discovered. 


1. Economic Aspects of Geology; by C. K Leith. 431 pages. Holt & Co., 
New York, 1921. 

The scientific and economic features of minerals, rocks, ores, and nonmetallic 
products. Written in nontechnical language. 

2. The Story of Copper; by Watson Davis. 380 pages. Century Co. , New York, 

A story of one metal, how it occurs, how it is extracted, and the uses to which it 
is put. Popularly written. 



Minerals compose the crust of the Earth and are therefore among the 
most common objects of daily observation. A mineral may be defined 
as a naturally occurring substance that has a definite chemical composi- 
tion and a definite combination of physical properties. This eliminates 
artificial products of the laboratory which may conform to the latter part 
of the definition. It also eliminates the natural products of organic 
agencies, as they do not show the definite chemical and physical char- 
acters of a mineral. 

Minerals are composed of chemical elements. Some consist of single 
elements, such as diamond and graphite (different forms of carbon), or 
gold and silver so far as they occur in the free state in nature; but most 
minerals are made up of two or more elements united in such a way as to 
give a product that differs in its properties from any of the elements 
composing it. 

At the present time about ninety different elements are recognized, 
but less than half of them are common and it has been calculated that 
more than 99 per cent of the crust of the Earth is composed of the follow- 
ing fourteen: 

Oxygen, 49.77 per cent 

Silicon, Si 26.09 

, Aluminum, Al 7.34 " 

Iron, Fe 4.11 

Calcium, Ca 3.19 " 

Magnesium, Mg 2.24 " 

Sodium, Na 2.33 

Potassium, K 2.28 " 

Hydrogen, H 0.95 " 

Titanium, Ti 0.39 " 

Carbon, C 0.18 " 

Chlorine, Cl 0.21 

Phosphorus, P 0.10 " 

Sulphur, S 0.10 " 

All others 0.72 

Total 100.00 



Three elements together with their symbols, not included in the above 
list, are added here because they are present in certain of the minerals to 
be studied. Each one composes but a small fraction of one per cent of 
the Earth's crust. 

Copper, Cu 

Lead, Pb 

Zinc, Zn 

The number of minerals formed by the combination of even these few 
elements is very great but the common ones are relatively few. 


It was said above that minerals have a definite chemical composition. 
This composition, as determined by analysis, serves to define and dis- 
tinguish the species, and indicates their relations to each other. Indi- 
vidual minerals react differently to various chemical reagents, and these 
reactions are one means of determining the kind of mineral under exam- 
ination. It is beyond the scope of this discussion to treat that aspect of 
mineralogy; but there are many text books that treat the subject fully. 


Structure of Minerals. Commonly the structure of minerals refers to 
their outward shape and form. The following descriptive terms are used 
in this connection, some of which are self-explanatory: crystallized, in 
definite crystals; columnar; fibrous; botryoidal, having a group of small 
rounded forms like a bunch of grapes; reniform (kidney-like) and 
mammillary, similar to botryoidal but in larger masses; foliated and 
micaceous, occurring in thin sheets; granular, in coarse to fine grains; 
corn-pad; earthy; stalactitic, formed in stalactite masses similar to icicles; 
massive, showing compact material with an irregular form; oolitic, formed 
of small, rounded grains which resemble fish roe, 

Crystals. The majority of minerals under favorable conditions will 
form in crystals. These are bodies which are bounded by plane surfaces 
that are arranged according to definite laws of symmetry. The division 
of mineralogy known as crystallography is important and interesting but 
one whose detailed study takes considerable time. Certain principles 
will be pointed out here however. Perfect crystals are exceptional things. 
The great majority of mineral specimens will not show them. When 
they are to be observed, however, they will help materially in the identi- 
fication of the mineral. In general the crystals of a certain mineral will 


show like or similar habits of crystallization. For instance, the mineral 
galena, PbS, characteristically crystallizes in cubes (Fig. 312). Magne- 
tite, Fe30 4 , commonly occurs in eight-sided crystals that are called octa- 
hedrons (Fig. 313). Garnets commonly occur in either dodecahedrons 
(Fig. 314) or trapezohedrons. These crystal habits are characteris- 
tic of these minerals and when recognized greatly aid in their identi- 

Cleavage and Fracture. The manner in which some minerals break 
is characteristic. If the break occurs with a smooth, plane surface the 
mineral is said to have a cleavage. This cleavage always takes place 
along planes, which may or may not be parallel to crystal faces. Some 
minerals will show but one cleavage, others two, three, or even six differ- 
ent cleavage planes. The number of planes of cleavage which a mineral 
shows and their relations to each other help to determine the mineral. 

312 313 314 315 

Fig. 312. Model of a cubic crystal. 

Fig. 313. Top of an octahedron showing four of the eight faces. 
Fig. 314. Showing six of the twelve faces of a dodecahedron. 
Fig. 315. A rhomb ohedr on. 

Good examples are the cubic cleavage of galena (in three planes at right 
angles to each other), the rhombohedral cleavage of calcite (three planes 
not at right angles so that the resulting form is rhombic; Fig. 315), the 
basal cleavage of mica (in one direction only). If a mineral does not 
show a cleavage it is said to have a fracture. Various kinds of fracture 
are as follows: conchoidal if the fractured surface is curved like the 
interior of a clam shell ; fibrous or splintery if it shows a fibrous character; 
uneven or irregular if the surfaces are rough. 

Color. The color of a mineral is one of its most conspicuous physical 
properties. The color of many minerals is a definite and constant prop- 
erty arid serves as an important means of identification. For example, 
the golden-yellow color of chalcopyrite, CuFeS 2 , the blue-gray of galena, 
PbS, the black of magnetite, Fe 3 4 , are striking properties of these 
minerals. However, surface alterations may change the color of a 
mineral, as is shown in the golden tarnish frequently observed on pyrite, 
FeS 2 . In noting the color of a mineral, therefore, a fresh surface should 


be examined. Moreover, many minerals show a variation in color in 
the different specimens. This may be due to a change in composition 
such as the gradual substitution of iron for zinc in the mineral sphalerite, 
ZnS, with the consequent darkening of the color of the mineral; or to 
impurities such as the red color given to quartz, Si0 2 , by the admixture 
of hematite, Fe 2 3 . Other minerals such as fluorite, CaF 2 , show a wide 
range in color without any apparent change in composition. 

Color of Powder or Streak. The color of the streak is an important 
aid to the identification of minerals. The streak is a thin layer of the 
powder of the mineral obtained by rubbing it upon an unglazed porcelain 
plate known as a streak plate. The color of the streak may be similar to 
the color of the mineral or quite different. For example, some varieties 
of hematite, Fe 2 3 , have a brilliant black color but give a red-brown 
streak that positively identifies the mineral. 

Luster. The luster of a mineral is the appearance of its surface due 
to the manner in which it reflects light. This must not be confused with 
color for two minerals with the same color may have totally different 
lusters just as a black paint with a shiny finish, such as an enamel, has 
an appearance different from that of a black paint with a dull finish be- 
cause it reflects light differently. 

Various descriptive terms are applied to the different kinds of luster 
exhibited by minerals. A partial list including the more important ones 
is given below: 

Metallic. Having the appearance of a metal. Example, pyrite, FeS 2 . 
(Most of the minerals with a black or dark-colored streak are included 

Vitreous. Having the luster of glass. Example, quartz, Si0 2 . 

Resinous. Having the appearance of resin. Example, sphalerite, 

Pearly. Having the iridescence of pearl. Example, some varieties of 
dolomite^ CaMg(C0 3 ) 2 . 

Greasy. Looking as if covered with a thin layer of oil. Some varieties 
of massive silica, Si0 2 . 

Silky. Like silk, as the result of a fine fibrous structure. Example, 
fibrous gypsum, CaS0 4 .2H 2 0. 

Adamantine. Having brilliant luster like that of a diamond, C. 

Hardness of Minerals. Minerals vary widely in their hardness, and 
a determination of this property is often an important aid to their 
identification. The relative hardness of any mineral may be told by 
comparing it with a series of minerals that has been chosen as a scale. 
The scale consists of crystallized varieties of the following minerals, each 
species being harder than those preceding it in the scale. 


Scale of Hardness 

1. Talc. 4. Pluorite. 8. Topaz. 

2. Gypsum. 5. Apatite. 9. Corundum. 

3. Calcite. 6. Orthoclase. 10. Diamond. 

7. Quartz. 

The relative hardness of any mineral in terms of this scale is deter- 
mined by finding which ones of these minerals it can and which it cannot 
scratch. In making the determination the following precautions must 
be observed. Sometimes when a mineral is softer than another, portions 
of the first will leave a mark on the second which may be mistaken for a 
scratch. It can be rubbed off, however, whereas a true scratch will be 
permanent. Some minerals are commonly altered on the surface to 
material which is much softer than the original mineral. A fresh surface 
of the specimen to be tested should therefore be used. Sometimes the 
physical structure of a mineral may prevent a correct determination of 
its hardness. For instance, if a mineral is pulverulent, granular or 
splintery in its structure, it may be broken down and apparently 
scratched by a mineral much softer than itself. It is always advisable 
when making the hardness test to confirm it by reversing the procedure, 
that is, by rubbing the unknown on the material of known hardness. 

The following materials will serve in addition to the above scale. The 
finger nail is a little over 2 in hardness, since it can scratch gypsum and 
not calcite.. A copper coin is about 3 in hardness, since it can scratch 
calcite. The steel of an ordinary pocketknife is just over 5 and ordinary 
glass has a hardness of 5.5. 

Specific Gravity. The specific gravity of a substance is stated as a 
number that indicates how many times heavier a given volume of the 
material is than an equal volume of water. Minerals show a range of 
specific gravity from about 1.5 to 20. The great majority of minerals 
range between 2.0 and 4.0. There are various instruments that enable 
one to determine the specific gravity of a mineral with more or less ac- 
curacy, but for ordinary purposes it is sufficient simply to judge the 
weight of a fair sized piece in the hand. After some practice rather small 
differences in specific gravity can be detected in this way and a mineral 
approximately located in respect to this property. 


Only a few of the more common minerals will be described. The stu- 
dent should always compare these descriptions with as many different 
specimens of the minerals as possible and should note the form, color, and 


luster of each sample and make the simple tests that determine the hard- 
ness, streak, and specific gravity. 


Composition. An oxide of iron, a combination of ferrous and ferric 
oxides, FeO.Fe 2 3 , or Fe 3 4 . 

Physical Characters. Color black. Streak black. Hardness 6. 
Heavy. Strongly magnetic. Usually granular or massive. Occurs in 
octahedral crystals (Fig. 313). 

Occurrence. An important iron ore. It is mined in New York 
State in the Adirondack Mts., in New Jersey and Pennsylvania, and 
many other parts of the world. It is common as a minor rock constituent, 
particularly in the darker colored igneous rocks. The black sand of the 
sea shore is largely magnetite. It sometimes occurs as a natural magnet, 
known as a lodestone. 


Composition. The ferric oxide of iron, FesOa. 

Physical Characters. Color reddish brown to black. Streak light 
to dark red-brown (Indian-red). Hardness 5.5-6.5. Commonly in 
botryoidal to reniform shapes with radiating structure. Often earthy. 
At times micaceous. Rarely in crystals. 

Occurrence. Hematite is a widely distributed mineral in rocks and 
forms the most abundant ore mineral of iron. More than nine-tenths of 
the iron produced in the United States comes from this mineral. The 
chief districts lie around the shores of Lake Superior in Michigan, Wis- 
consin, and Minnesota. Important districts are also located in northern 
Alabama and eastern Tennessee. Hematite forms the cementing mate- 
rial in red sandstone. It is used also in red paints and as a polishing 
material, known as rouge. 


Composition. Hydrous ferric oxide, Fe 2 3 .H 2 0. 

Physical Characters. Color dark-brown to nearly black. Streak 
yellowish brown. Hardness 5-5.5. Medium heavy. Commonly in 
mammillary to stalactitic forms with radiating fibrous structure; some- 
times earthy. * 

Occurrence. A minor source of iron. Limonite is a common mineral 
formed through the alteration or solution of previously existing minerals 
containing iron. It is found as a cellular mass known as gossan in the 
upper part of sulfide veins; as loose, porous bog-iron ore; associated with 
siderite as large deposits in limestone and other rocks. It gives the 


brown and yellow color to many weathered rocks, sedimentary strata, 
and soils. 


Composition. Iron sulphide, FeS 2 . 

Physical Characters. Color pale brass-yellow. Streak black. Hard- 
ness 6-6.5 (unusually hard for a sulphide). Heavy. Usually granular. 
Sometimes in crystals, commonly striated cubes or octahedrons. 

Occurrence. The most common sulphide mineral. Found in many 
rocks and is an important vein mineral. Often carries small amounts of 
gold or copper and so becomes an ore for both these metals. Never 
serves as an ore of iron but is used as a source of sulphur in the manu- 
facture of sulphuric acid. Its presence in building stones detracts from 
their value since by its oxidation sulphuric acid is formed, which causes 
disintegration of the rock and iron oxide which stains its surface. 


Composition. Copper-iron sulphide, CuFeS 2 . 

Physical Characters. Color golden-yellow; often tarnished to bronze 
or iridescent colors. Streak greenish black. Hardness 3.5 (note differ- 
ence from pyrite). Heavy. Usually compact massive, rarely in crys- 

Occurrence. A common and important ore mineral of copper. Oc- 
curs widely distributed in vein deposits with many other sulphide 


Composition. Zinc sulphide, ZnS. Nearly always contains a small 
amount of iron. 

Physical Characters. Commonly yellow brown to dark brown in 
color. Darkens with increase of iron. Resinous to submetallic luster. 
Hardness 3.5-4. Heavy. White to yellow and brown streak, always a 
lighter shade than the mineral itself. Perfect dodecahedral cleavage. 
Usually massive cleavable. 

Occurrence. The most common and important source of zinc. 
Found widely distributed but generally in veins or irregular bodies in 
limestone. Often associated with galena, pyrite, and chalcopyrite. 


Composition. Lead sulphide, PbS. 

Physical Characters. Color lead-gray. Streak grayish black. Hard- 
ness 2.5 (soft). Very heavy. Bright metallic luster. Perfect cleavage 


in three planes at right angles to each other, forming cubes. Often also 
in natural cubic crystals (Fig. 312). 

Occurrence. The most common and important source of lead. 
Frequently associated with silver and often serves as an ore of that metal. 
Also commonly found with zinc minerals. 


Composition. Calcium carbonate, CaC0 3 . 

Physical Characters. Color usually white or colorless. May be 
variously tinted, gray, red, green, blue, etc. Usually transparent to 
translucent. Hardness 3. Light in weight. Perfect cleavage in three 
planes at oblique angles to each other, giving rhombic-shaped faces 
(rhombohedral cleavage) (Fig. 315). Often in crystals which generally 
have a hexagonal cross section. Will effervesce freely on application of a 
drop of cold acid. This will serve to distinguish calcite from dolomite, 
CaMg(COg)2, another common carbonate, which will not show efferves- 
cence under these conditions. 

Occurrence. A very common mineral. Is the chief constituent of 
limestones and marbles. Also a very common vein mineral. Used for 
the production of lime, plasters, and cement. 


Composition. Carbonate of calcium and magnesium, CaMg(COs)2. 

Physical Characters. Usually white or gray. Sometimes flesh-col- 
ored. Transparent to translucent. Hardness 3.5-4 (harder than 
calcite). Perfect cleavage in three planes not at right angles to each 
other (rhombohedral cleavage). Light in weight. Vitreous to pearly 
luster. Will not effervesce upon application of a drop of cold acid unless 
the specimen is scratched or powdered (differs from calcite). Found in 
cleavable masses and in crystals which sometimes have curved faces. 

Occurrence. Composes rock masses such as dolomite limestone and 
marble. Also as a vein mineral. Often intimately mixed with calcite. 
In the rock form, used as a building and ornamental stone, for the manu- 
facture of some cement, and as a source of magnesia for refractory sub- 


Composition. Hydrous calcium sulphate, CaS0 4 .2H 2 0. 
Physical Characters. Usually white or colorless. Hardness 2 
(easily scratched with the finger nail). Light in weight. Has one very 


perfect cleavage. May be in tabular diamond-shaped crystals or in 
cleavable masses. Often also fine granular, sometimes fibrous. 

Occurrence. Is a common mineral which is widely distributed in 
sedimentary rocks, often in thick beds. It frequently occurs inter- 
stratified with limestones and shales. Often found in connection with 
salt beds. Forms twinned crystals similar to Fig. 316. Is. chiefly used 
for the production of plaster of Paris. 


Composition. Sodium chloride, NaCl. 

Physical Characters. Usually white or colorless. Hardness 2.5. 
Light in weight. Perfect cleavage in three planes at right angles to each 
other (cubic cleavage). Transparent to translucent. Salty taste. 
Generally in cubic crystals or in masses showing cubic cleavage. 

Occurrence. In extensive beds or irregular masses interstratified 
with sedimentary rocks and associated with gypsum. Used for culinary 
and preservative purposes; also very extensively in chemical industry. 


Composition. Silicon dioxide, SiOg. 

Physical Characters. Usually colorless or white, but frequently 
colored by different impurities, yellow, red, pink, amethyst, green, blue, 
brown, black. Vitreous luster. Transparent to opaque. Hardness 7. 

316 317 318 

Fig. 316. Model of a twinned crystal of gypsum. 

Fig. 317. Model of a quartz crystal. This is a six-sided prism terminated by two 
unequally developed rhombohedrons. 

Fig. 318. Model of an orthoclase crystal. 

Light weight. Conchoidal fracture. Commonly in hexagonal* crystals 
similar to Fig. 317. The triangular faces at the ends of the crystals 
are usually smooth while the rectangular faces about the middle of 
the crystals are horizontally striated. Also massive. 

Varieties. There are many varieties of quarts to which different 
names are given. A few are as follows: Rock crystal, colorless quartz, 


commonly in distinct crystals; Amethyst, quartz colored purple or violet; 
Rose Quartz, usually massive with a pink color; Smoky Quartz, quartz 
with a smoky yellow to brown or almost black color; Chalcedony, finely 
fibrous material, translucent with a waxy luster; Agate, a variegated 
chalcedony often delicately banded with different colors; Jasper, ex- 
tremely fine grained quartz colored red with hematite. 

Occurrence. Quartz is one of the most common minerals. It occurs 
as an important constituent in many rocks. It is also the most com- 
mon vein mineral It makes up the largest part of sands and sandy 
soils. It is widely used in its various colored forms as ornamental 
material. It is used for abrading purposes, in the manufacture of glass, 
porcelain, in paints, scouring soaps, etc. As sand it is used in mortars 
and cements. Quartzite and sandstone, rocks made up largely of quartz, 
are used in building, etc. 


Composition. There are several different garnets which vary from 
each other in the elements they contain. They are all silicates with 
somewhat similar formulas. The most common garnet contains ferrous 
iron and aluminum, FesA^SiO^s. Other garnets contain magnesium, 
calcium, manganese, ferric iron, chromium, etc. 

Physical Characters. Color varies with the composition. Most 
commonly red or brown. May be yellow, white, green, black. Trans- 
parent to almost opaque. Hardness 7. Medium heavy. Usually 
distinctly crystallized, either in a form showing twelve rhombic-shaped 
faces (dodecahedron, Fig. 314) or twenty-four trapezium-shaped faces 

Occurrence. Garnet is a common and widely distributed mineral, 
occurring as an accessory mineral in various kinds of rocks. Used as an 
inexpensive gem stone and because of its hardness as an abrasive material. 


Composition. Potassium aluminum silicate, KAlSi 3 8 . 

Physical Characters. Colorless, white, gray, flesh-red, more rarely 
green. Streak white. Hardness 6. Light in weight. Has two good 
cleavages making 90-degree angles with each other (whence name of min- 
eral). Sometimes in crystals, usually as in Fig. 318. 

Occurrence. The most common' silicate. Widely distributed as a 
prominent rock constituent, occurring in many kinds of rocks. Also in 
large crystals and cleavage masses in what are known as pegmatite veins. 
From these veins it is quarried in large amounts for use in the manufac- 
ture of porcelain. 



Composition. Sodium-calcium aluminum silicates. 

Physical Characters. Various shades of gray, less commonly white. 
Transparent to opaque. Hardness 6. Light in weight. Have two 
cleavages making nearly a 90-degree angle with each other. Commonly 
distinguished from orthoclase by the color, by the presence on cleavage 
surfaces of a series of fine parallel striation lines, or by a bluish opales- 
cence. Sometimes crystallize in thin bladed crystals with a curved 
surface and a pearly luster. 

Occurrence. In much the same manner as orthoclase. 


Composition. A complex silicate containing potassium and alu- 

Physical Characters. Possesses a perfect cleavage in one direction 
which allows the mineral to be split into excessively thin sheets. The 
folia are flexible and elastic. Transparent and almost colorless in thin 
sheets. In thicker blocks, opaque with light shades of brown and green. 
Hardness 2-2,5. Light in weight. Structure foliated in large to small 
sheets, sometimes in scales. 

Occurrence. A common rock-making mineral. It is found in granite 
together with quartz and a feldspar and, with the same associations, it 
occurs in pegmatite veins. Characteristic of a series of rocks made up 
largely of mica, the minerals being arranged in parallel layers so that the 
rocks possess a cleavage. These rocks are known as mica schists. Is 
used chiefly as an insulating material in the manufacture of electrical 
apparatus. Used as a transparent material in stove doors, etc. There 
are many other minor uses. 


Composition. A complex silicate containing potassium, magnesium, 
and aluminum. 

Physical Characters. Perfect micaceous cleavage. Folia flexible 
and elastic. Color usually dark-green and brown to black. Thin 
sheets usually have a smoky color (differing from the almost colorless 
muscovite). Hardness 2.5-3. Light in weight. 

Occurrence. An important and common rock-making mineral but 
not as common as muscovite. 



Composition. Complex silicate containing magnesium and aluminum. 
A complex group of minerals of similar characters which are called col- 
lectively the Chlorites from their common green color. 

Physical Characters. Perfect micaceous cleavage. Folia flexible but 
not elastic (differing from muscovite and biotite). Color green of various 
shades. Hardness 2-2.5. Light in weight. 

Occurrence. A common rock-making mineral, usually of secondary 
origin. It results from the alteration of silicates containing aluminum, 
ferrous iron, and magnesium. To be found where rocks containing such 
minerals have undergone considerable change due to the heat and pres- 
sure to which they have been subjected and have become metamorphic 
rocks. The green color of many rocks is due to the presence of this 
mineral. This is particularly true of many schists and slates. 


Composition. A magnesium silicate, 

Physical Characters. Olive-green or yellow-green to blackish-green. 
Luster greasy or wax-like, silky when fibrous. Hardness 2.5 to 5, usually 
4. Light weight. Usually massive but also fibrous or felted. 

Occurrence. A common mineral and widely distributed. Always as 
an alteration product of some magnesian silicate. The massive variety 
is sometimes used as an ornamental stone. The fibrous variety known 
as chrysotile is the chief source of asbestos. 


These two common and important rock-making minerals are similar 
in some respects, and consequently are difficult to distinguish from one 
another in ordinary rock masses, where good crystal forms do not occur. 
However it is well to study them separately under favorable conditions, 
in order to appreciate their differences as well as their points of similarity. 


1 Composition. A silicate containing chiefly calcium and magnesium; 
also varying amounts of aluminum, iron, sodium, etc. 

Physical Characters. Color usually from light to dark green varying 
with amount of iron. Also at times nearly white or black. Transparent 
to opaque. Hardness 5-6. Light in weight. Often in prismatic crys- 
tals with eight sides (Figs. 319 and 320). The angle between alternate 
faces is nearly 90 degrees. These faces will fit into the corner of a box or 


tray. ^ It is by these angles that the mineral can best be told from 
amphibole. Some specimens show a fair cleavage parallel to the faces 
lettered m in the figures, the angle between the cleavage faces being also 
nearly 90 degrees. 

Occurrence. Pyroxene is a common and important rock-making 
mineral, being found chiefly in the dark-colored igneous rocks. Seldom 
found in rocks that contain much quartz. 


Composition. Silicate of calcium and magnesium with varying 
amounts of aluminum, iron, sodium, etc. Similar to pyroxene. 

Physical Characters. Color usually light to dark green varying with 
amount of iron. Also nearly white or black. Transparent to opaque. 
Hardness 5-6. Light in weight. Often in prismatic crystals with six 
sides (Figs. 321 and 322). Figure 321 shows that the angles between the 

Fig. 319. Cross section of pyroxene normal to the long axis of the crystal. The 
cleavage traces are parallel to the prism faces (marked m) . The alternate faces (those 
marked m or those unmarked) make angles of approximately 90 with each other. The 
cross section has eight sides. 

Fig. 320. Model of pyroxene crystal. Shows the prism faces (m) in perspective and 
a cross section at right angles to the prism zone similar to that in Fig. 319. The inclination 
of the model somewhat distorts the interfacial angles. 

Fig. 321. Cross section of amphibole normal to the long axis of the crystal. The 
cleavage traces are parallel to the prism faces (m) which make angles of 124 and 56 
with each other. The cross section has six sides. 

Fig. 322. Model of an amphibole crystal. Shows the prism faces (m) in perspective 
and a crows section at right angles to the prism zone similar to that in Fig. 321. The in- 
clination of the model somewhat distorts the interfacial angles. 

faces lettered m are 124 and 56 (very different from the corresponding 
angles in pyroxene). Has a good cleavage parallel to the faces lettered 
m. The differences between the crystals and the cleavage angles in 
pyroxene and amphibole, and the fact that amphibole has the better 
cleavage constitute the chief distinctions between the two. Amphibole 
usually has a higher luster and a smoother surface than pyroxene. In 
some varieties of amphibole the crystals are long and needlelike, resulting 
in a fibrous structure. Pyroxene does not occur in this form. 

Occurrence. Amphibole is an important rock-making mineral oc- 
curring in both igneous and metamorphic rocks, being particularly 


characteristic of the latter. Often recognized in the rocks by its elon- 
gated bladed structure and good cleavage. 

Hornblende is a common dark variety of amphibole. 

Pyroxene and amphibole together with biotite are the common dark 
constituents of nearly all crystalline rocks. The first two can be sepa- 
rated from biotite by the fact that they occur in prismatic crystals that 
cannot be divided into thin folia; that is, they lack the perfect basal 
cleavage of the micas. When present as small grains in a rock they lack 
the high luster characteristic of small brilliant flakes of biotite. They 
can be told from chlorite by their much greater hardness as well as their 
lorm and lack of foliation.